NOVEMBER 2010

HPIMPACT SPECIALREPORT BONUSREPORT

LCFS impact oil sands PLANT SAFETY AND HEAT TRANSFER

refiners ENVIRONMENT New designs foster
US demand for Methods to mitigate efficiency and uptime
lubricants increasing CO2 and other air
emissions

www.HydrocarbonProcessing.com
 NOVEMBER 2010 • VOL. 89 NO. 11
www.HydrocarbonProcessing.com

SPECIAL REPORT: PLANT SAFETY AND ENVIRONMENT

31 Stretch in technology and gaps in process safety

for the hydrocarbon industry Cover The Kårstø gas processing plant on
the west-coast of Norway is the largest
Industry faces challenges to prevent and control hazards of its type in Europe. The installation
plays a key role in the transportation
M. S. Mannan and treatment of gas and condensate
(light oil) from important areas on the

33 Better water management:

A crucial and growing requirement
Norwegian continental shelf. Thirty fields
are linked to Kårstø via pipelines. Every
day millions of cubic meters of gas and
Technologies for water reuse and recycling needed now non-stabilized condensate flow into the
plant, where the heavier components are
G. Messina removed by separation. The remainder,
known as dry or sales gas, is exported by
35 Water among causes for storage tank explosion
Reinvestigation uncovers true accident events
pipeline to the Continent.

M. Ferjencik and B. Janovsky

41 Consider new analysis for flares

Applying dynamic models in designing safety systems can reduce capital costs
Z. Urban, M. Matzopoulos, J. Marriott and B. Marshall HPIMPACT
17 LCFS will adversely
49 Designing the correct pressure-relieving system
Use these relief rate calculations for gas thermal expansion as a cause for overpressure
affect oil-sands
crude refiners
S. Rahimi Mofrad
17 US demand for

55 Customize operator training for your thermal oxidizers

This case history shows the benefits of site-specific programs
lubricants to reach 2.25
billion gallons in 2014
in new equipment installations 17 Linde supplying
T. Gilder, D. Campbell, T. Robertson and C. Baukal hydrogen technology
to US BMW plant
61 Emergency response planning—
start at the plant design stage 18 Oil and gas reserves
increased 3 percent
Follow these guidelines for a safer facility
in 2009
R. Saini

65 Optimized fired heater control

Residual oxygen measurement principle lowers emissions and improves efficiency
A. J. Mouris
COLUMNS
69 Consider real-gas modeling for turboexpanders
New visualization methods expose problems with traditional designs 9 HPIN RELIABILITY
K. Kaupert Wire mesh vs. wire
size for temporary

73 Predictive emissions monitoring

helps reduce stack air emissions
strainers
11 HPIN EUROPE
New technology reduces compliance costs while optimizing operations Consultants stand
R. Hovan by need for
swingeing cuts
79 Consider switching to Internet protocol surveillance
Here’s a checklist to make the jump
13 HPINTEGRATION
STRATEGIES
M. S. Wilson Future of the
collaborative process
automation system
ENGINEERING CASE HISTORIES
15 HPI VIEWPOINT
81 Case 59: Heat-up rates and thermal cracking
A good analysis is usually better than speculation
Weighing on dual-
temperature control
T. Sofronas
86 HPIN WATER
MANAGEMENT
DEPARTMENTS Utility water boot
7 HPIN BRIEF • 23 HPINNOVATIONS • 27 HPIN CONSTRUCTION camp for process
29 HPI CONSTRUCTION BOXSCORE UPDATE engineers—Part 3
82 HPI MARKETPLACE • 85 ADVERTISER INDEX
Select 77 at www.HydrocarbonProcessing.com/RS
 HPIN BRIEF
BILLY THINNES, NEWS EDITOR

BT@HydrocarbonProcessing.com

The US Environmental Protection Agency (EPA) has waived a ■ Better gas

limitation on selling fuel that is more than 10% ethanol for model year 2007 and
newer cars and light trucks. The waiver applies to fuel that contains up to 15% etha- storage via MOFs
nol—known as E15—and only to model year 2007 and newer cars and light trucks.
This represents the first of a number of actions that are needed from federal, state and Natural gas-powered vehicles may
soon be able to travel double the dis-
industry towards commercialization of E15 gasoline blends. A decision on the use of
tance on a single tank—due to metal
E15 in model year 2001 to 2006 vehicles will be made after EPA receives the results organic frameworks (MOFs). BASF
of additional Department of Energy testing, which is expected to be completed in research scientists have developed an
November. However, no waiver is being granted this year for E15 use in model year innovative method for the solvent-
2000 and older cars and light trucks—or in any motorcycles, heavy-duty vehicles or free industrial-scale manufacture of
non-road engines—because currently there is no testing data to support such a waiver. those materials for better gas storage.
Since 1979, up to 10% ethanol or E10 has been used for all conventional cars and MOFs produced by the new method
light trucks, and non-road vehicles. are currently being trialed for natural
gas storage in heavy-duty vehicles.
With their special structure and
GE has snapped up Dresser for a cool $3 billion. The deal includes large surface area, MOFs open up new
all of the Dresser businesses that provide products and services for compression, flow opportunities for alternative propul-
technology, measurement and distribution infrastructure for customers in more than sion systems, in catalysis, as nano-
150 countries. Dresser’s extensive, global-installed base of products generate aftermar- reactors, and in drug delivery, mak-
ket service revenues in excess of 40% of total revenues. Morgan Stanley acted as the ing them hugely interesting both for
exclusive financial advisor to Dresser on this transaction. industry and university research.
“This substance class opens up
new areas of applications in mate-
Strong Petrochemical Holdings Ltd. has entered into an exclusive
rial science. We are delighted at this
marketing agreement with Eurocontrol Technics Inc. Eurocontrol will appoint Strong significant advance in industrial-
as the exclusive sole agent to market and sell Eurocontol’s Petromark fuel authentica- scale production, which is a crucial
tion system to potential customers in China, Hong Kong and other defined target requirement for the commercial use
markets throughout Asia. This exclusive agreement is contracted for an initial period of these fascinating materials,” said
of three years and is renewable for further periods of three years. Strong will act as the Dr. Friedrich Seitz, head of research
exclusive sole agent, responsible for the marketing and sale of the products in the terri- chemicals at BASF.
tories and Global Fluids International S.A., a wholly owned subsidiary of Eurocontrol, BASF has been working toward
industrial-scale synthesis of metal-
will be responsible for the design and technical support to the consortium. Strong will
organic frameworks for the past 10
bring to the consortium its extensive business connections in the target markets with years. MOFs are highly crystalline struc-
participants in the petrochemical complex. tures with nanometer-sized pores that
allow them to store hydrogen and
Petrofac has a contract with the government of Sharjah, UAE, to other high-energy gases. The larger
take over operational responsibility and facilities management of the Sajaa gas plant specific surface area and high porosity
and related assets, located approximately 30 km from Sharjah, UAE. The five-year on the nanometer scale enable MOFs
contract was awarded following a competitive open bidding process and is worth to hold relatively large amounts of
in excess of $250 million. The government of Sharjah, acting through the Sharjah these gases. The pores are adjustable
in terms of size and polarity and so can
Petroleum Council, holds a 60% participating interest in the Sharjah gas and associ- be fine-tuned for specific applications.
ated liquids concession. Used as storage materials in the
natural gas tanks of municipal util-
Marathon Petroleum Co. LP (MPC) is selling most of its downstream ity vehicles, MOFs offer a docking
assets in Minnesota to ACON Investments, LLC, and TPG Capital, LP. ACON and area for gas molecules, which can
TPG formed Northern Tier Energy LLC to operate the assets as a stand-alone com- be stored in higher densities as a
pany. Included in the transaction will be the 74,000 bpd St. Paul Park refinery and result. The larger gas quantity in the
associated terminals, 166 SuperAmerica convenience stores (including six stores in tank increases the vehicle’s range. An
advantage of the production method
Wisconsin), SuperMom’s LLC, SuperAmerica Franchising LLC, interests in pipeline developed by BASF is that it uses no
assets in Minnesota and associated inventories. The total sales value is approximately organic solvents. The simple method
$900 million, including Northern Tier preferred stock with a stated value of $80 mil- gives a higher material yield from an
lion. Approximately $300 million of the total sales value is for the inventories associ- aqueous medium and is suitable for
ated with these operations. The agreement also contains earnout and margin support existing BASF production plants.
components where Marathon could receive up to an additional $125 million over eight MOFs were discovered toward the
years or may be required to provide up to $60 million of margin support to the buyers, end of the 1990s by US chemist Omar
subject to certain conditions. HP M. Yaghi. HP

HYDROCARBON PROCESSING NOVEMBER 2010

I7
 HPIN RELIABILITY
HEINZ P. BLOCH, RELIABILITY/EQUIPMENT EDITOR

HB@HydrocarbonProcessing.com

Wire mesh vs. wire size for temporary strainers

Even an experienced reliability professional may find some wire diameter equals 0.333 in.). The tables also tell us that we have
customary terminology dealing with wire gauge and wire mesh ~35% of the area open for liquid flow and ~65% of the area would
to be confusing. Our HPIn Reliability column of April 2009 be taken up by the 0.135-in. diameter guard screen wires.
mentioned pump suction strainers made of 20-gauge wire mesh Next, one would determine the wire mesh utilizing the
and a few words about wire gauge will be helpful. 20-gauge wires. That particular wire cloth, to be placed inside
Wire gauge (also known under the alternative spelling “wire the three-mesh guard screen, should use 20-gauge wire. Reference
gage”) is indicative of electrical wire diameter, with primary impli- 2 shows the related wire diameter (0.032 in.) in the middle of a
cations for allowable electric-current flow (amperes). According six-mesh table. The openings are 0.1347 in., and adding the two
to working tables published for Standard Annealed Solid Copper numbers gives us 0.1667 in.—six squares per in.
Wire,1 20-gauge wire has a diameter of 0.032 in. and, at about Using the above information and Fig. 1 will equip us to ask
40°F, has a resistance of approximately 11 ohms per 1,000 ft of a draftsman to design a pointed cone—the three-mesh guard
length. While this month’s column has nothing to do with electric screen—with a surface area about three times that of the pipe or
current, we want to assume that the 20-gauge stainless-steel wire spool piece cross-sectional area. The six-mesh metal cloth would go
we would like to see inside the three-mesh guard screen would be on the inside of the cone and only then could we call it a strainer.
woven from wire that has a diameter of 0.032 in. On any permanent strainer installation, it would be wise
The terms wire mesh and wire cloth are often used interchange- to monitor the delta-p (the pressure drop) between the mesh
ably. They denote a metal wire weave with its various implications upstream and downstream sides. Clogged strainers can cause seri-
regarding available flow-through area as a percentage of the total ous machine malfunction and costly damage. HP
square inches of wire cloth area. Our April 2009 column meant
LITERATURE CITED
to recommend placing a wire mesh made from 20-gauge wire inside 1 Marks’ Standard Handbook for Mechanical Engineers, 7th Edition (1969),
a three-mesh guard screen. McGraw-Hill Book Company, New York, New York.
So, to clarify: While wire mesh made from 1⁄32-in. wire is to be 2 www.mcnichols.com/products/wiremesh.
placed inside a three-mesh guard screen, no 20-mesh wire cloth is
involved. The three-mesh guard screen is usually recommended The author is Hydrocarbon Processing’s Equipment/Reliability Editor. A practic-
for temporary strainers to reinforce a finer mesh. ing consulting engineer with close to 50 years of applicable experience, he advises
process plants worldwide on failure analysis, reliability improvement and main-
• “Three-mesh” means wire cloth (or guard screen in the illus- tenance cost-avoidance topics. He has authored or coauthored 17 textbooks on
tration published in April 2009). On a piece of paper, that would machinery reliability improvement and over 470 papers or articles.
be three wires per in. in the x-direction (left-to-right) and three
wires-per-in. in the y-direction (up-and-down). Similarly, eight
Identification tab at
mesh would be eight wires in the x-direction and eight wires in Strainer screen
top for raised-face
Spool piece
the y-direction; or 200 mesh would be 200 wires in the x-direction Screen seam
flanges. Mount on
flange bolts.
and 200 wires in the y-direction, etc.
Preferred flow
• In a three-mesh screen there are, therefore, 3 x 3 = 9 openings. direction
• In a three-mesh screen, the distance from the center of one
wire to the center of an adjacent wire is 1⁄3 in., or 0.333 in.
• One manufacturer2 of wire cloth offers three-mesh cloth Strainer assembly
made with wires ranging from 0.031 in. to 0.162 in.—in each case, Install gasket on each
side of strainer flange.
there would be nine openings and the distance from the center of a 12-gauge x 1-in. wide
identification tab.
wire to the center of an adjacent wire would always be 0.333 in. Mount mesh made from ½-in.
20-gauge wire inside
• If one were to pick the 0.162-in. thick wire, one would have a three-mesh guard ½-in. 12-gauge
ASTM “A”
greater strength than if one picked the 0.031-in. wire. screen and stagger the 167 type
longitudinal seams.
• Likewise, if one picked the 0.162-in. wire diameter, the 316
90° 1-in.
resulting open area would only be 26%; whereas, if one selected Detail “A”
the 0.03-in. wire diameter, there would be an 82% open area. Both sides of strainer
• For temporary strainers in process pump piping, one would Screen section flange must be free of
gouges, weld spatter
pick a three-mesh guard screen with a wire diameter of about See detail “A” or other imperfections
0.135 in. To reiterate, the wire-to-wire center distance would be that might impair proper
gasket sealing.
0.333 in. and there would be nine openings. The wire-cloth manu-
facturer’s tables (obtained from the Internet2) for this diameter give FIG. 1 Temporary strainer for pump suction piping.
an opening width of 0.198 in. (remember that opening width plus
HYDROCARBON PROCESSING NOVEMBER 2010
I9

TIM LLOYD WRIGHT, EUROPEAN EDITOR
tim.wright@gulfpub.com
Consultants stand by need for massive capacity cuts
In November 2009, I wrote about how the world’s down-
stream consultants seemed almost to be vying to ring up the
most pessimistic prognosis for refinery closures. A year on,
there’s little to suggest they were off the mark. Talking to one of
them today, they’re just as bearish as they were then. You may
recall that JBC Energy in Vienna told me last year that 2.8 mil-
lion bpd (MMbpd) of European refinery capacity must close.
By February of this year, they increased that estimate by nearly
600,000 bd.
How much to cut? From its work for OPEC, the US consult-
ing group, Ensys said that 10 MM bpd of cuts would be required
to bring global utilization rates up to 85% by 2020. Energy Mar-
ket Consultants (EMC) in London, part of Facts Global Energy,
had perhaps the most dire warning, putting the level of necessary
reductions at 11 MMbpd by 2020.
This was when I asked Leif Nilsson, the canny trading man-
ager at Preem, by then specializing in storing up oil for better
times ahead, for his advice for refiners. “Margins are going to be
bleak,” he said. “Just try and sit it out.”
Margins are still bleak. Margins have come back some
since the depths of the economic crash last year, but they’re still
way below the pre-crash glory days. EMC’s variable cost FCC
margin is hovering around $1.50/bbl, and, with some cuts of the
barrel, it is heavily negative. Refiners have clearly concluded that
they need to take action to reduce their exposure.
Shell has announced the closure of one site, albeit at the far
extent of the Atlantic Basin in Montreal. It has successfully con-
cluded the sale of a German site, Heide, to a private investor, and
is said now to be in exclusive talks to sell its Gothenburg, Sweden
refinery to a Finnish company with fuel retailing interests.
In its most recent results presentation to analysts, Total
reiterated its plans to reduce its refining output by 500,000
bpd by 2011. There was, however, no new information on its
long-standing effort to sell the UK Lindsay refinery, or its UK
retail network. Meanwhile, Total’s abortive attempt to negotiate
redundancies and the closure of its Dunkirk, France, refinery are
thought to be what has given some potential investors in Europe’s
downstream industry pause for thought.
I hear that Essar of India’s team, which has been looking
over key assets like the 246,000-bpd Shell Stanlow refinery, got
itchy feet after watching, as on June 30, a French court ordered
Total to restart its Dunkirk refinery within 15 days, or face a
€100,000 fine.
Hope floats. There are some small signs of light at the end
of this dark tunnel. For one thing, a key distillate market is in
flux. Heating oil futures have been in contango since the massive
HPIN EUROPE
demand destruction that accompanied the 2008 money supply
crisis. With no buyers now, futures markets have consistently
reflected premiums for oil delivered in the future. It’s a bearish,
rather than optimistic, signal, suggesting a belief that things can
only get better because they’re so bad now.
In just the last few days, as I write, that structure has flipped
around, although just for the coming months before winter. Sud-
denly and admittedly compared to a very low baseline, Europe
is the top destination for surplus cargoes of distillate. A surge
of opportunistic vessel fixtures this week indicated that US and
Asian traders and refiners, as predicted in my October column
incidentally, are attempting to capture some of a $17/metric ton
spread between European ULSD/50-ppm heating oil and spot
prices in the US.
Reflecting the improved demand, the OPIS Spot Northwest
Europe Gasoil Cargoes to Brent Crude crack averaged $11.24/
bbl for the first three weeks of September this year, compared to
$5.74/bbl for the same three weeks a year earlier.
Still too much spare capacity. So when I spoke to Roy
Jordan at EMC this week, I asked if he thought the reversed
market structure was indeed a sign that things are on the up for
European refiners. “Unfortunately, you’ve got to see it in the con-
text of a structural surplus of gasoline and an inability to export
to the US,” he said. “Yes, that heating oil crack is improved, but
you have to see it against very weak fuel, gasoline and naphtha
cracks.” Jordan says that the heating oil market, itself driven to a
great extent by diesel demand, is Europe’s essential oil. Take that
profitable cut away from European refiners, and they might as
well shut up shop altogether, he suggests.
EMC’s view of the recent backwardation? “We’ve got it
down to the refinery maintenance season,” he says. And so, as
EMC prepares new figures for a forthcoming refining outlook,
I asked Jordan if EMC is sticking to its guns on the need for
cuts. When I last spoke to EMC, they were estimate a need for
7 MMbpd of reductions by 2015, with a further 4 MMbpd
necessary by 2020.
“There’s no retreat since then,” says Jordan. “We’re more con-
firmed than ever. The spare capacity is enormous and margins
are not recovering.” HP
The author is HP’s European Editor and also a specialist in European distillate
markets. He has been active as a reporter and conference chair in the European
downstream industry since 1997, before which he was a feature writer and reporter
for the UK broadsheet press and BBC radio. Mr. Wright lives in Sweden and is the
founder of a local climate and sustainability initiative.
HYDROCARBON PROCESSING NOVEMBER 2010
I 11
 “Advertisement”

ReactoGard® V
The Ultimate Protection
for Fixed Bed Catalytic
Reactors from Eaton.
Using refinery-proven
AccuFlux™ media,
ReactoGard V automated
backwashing systems offer
significant advantages in all
types of refinery feedstocks.
ReactoGard V systems are engineered to
operate cleaner, more efficient low flux
sizing. This is achieved through improved
media materials and AccuFlux element
designs that increase filtration area by as
much as 300% over previous standards. In
addition, a new multi-bank system and con-
trols approach eliminates the headers and
reduces the number of valves by up to 85%.
The ReactoGard V provides:
• Improved filter productivity
•
•
Complete media cleaning
Superior mechanical integrity for enhanced
serviceability and reduced maintenance
• Reduced product loss through less
frequent backwashing
• Reduced cost of installed filtration capacity
• Reduced space requirements
For more information, call Eaton
at 888-309-6648.
Select 118 at www.HydrocarbonProcessing.com/RS
ClearAmineTM
A Safe, Green and Economical Solution from
Eaton for the Filtration of Amine Gasses.
With the ClearAmine™ automated self-cleaning
system there are no bags or cartridges to purchase,
no material to inventory and track, no hazardous
waste landfill fees and greatly reduced maintenance
and handling costs. Over a ten year period, you
can expect nearly a 60% savings compared to
a disposable media system. Plus, with no cartridges or bags to change out, employee
exposure to hydrogen sulfide is minimized or eliminated. Amine filtration is a key factor
for trouble-free sulfur removal. Pipe scale, corrosion, iron sulfide, sand and other solids
build-up in closed-loop amine systems. They foul flash drums, heat exchangers, strippers,
re-boilers and carbon filters and also attract hydrocarbons – which cause system foaming.
ClearAmine™ provides an efficient method of removing harmful solids to protect your
investment, extend equipment life, and maximize your amine system’s effectiveness.
For more information, call Eaton at 888-309-6648.
Select 119 at www.HydrocarbonProcessing.com/RS
Your complete source for
Oil & Gas Liquid Filtration
• ReactoGard® V
• ClearAmineTM
• Strainers
• Bags & Housings

Model 2596
Automatic Self-
Cleaning Strainer
This automatic
self-cleaning
strainer is designed
for the continuous
removal of entrained Sustainable. Economical. Environmental.
solids from liquids in Eaton provides complete refinery and petrochemical liquid filtration solutions
pipeline and process with installations in thousands of applications worldwide. We stand behind
water systems.
our products with comprehensive parts and service programs, plus innovative
With an auto control system monitoring the
purchase/lease options. Count on Eaton to provide complete solutions that are
strainer operation, cleaning is accomplished
sustainable economically and environmentally while protecting the safety of
by an integral backwash system. A small
your workforce. At almost any point in your refinery, effective, efficient filtration
portion of the screen element is isolated
can deliver tremendous benefits including:
and cleaned by reverse flow. The remaining
screen area continues to strain – providing
„ Longer equipment life „ Protection of costly process materials
an uninterrupted process. Also included is
„ Reduced downtime „ Extended runtime between turnarounds
the idLTM shaft seal that positively prevents Enhanced environmental stewardship
„ Improved safety „
leakage from the backwash shaft at the top
of the strainer. Eaton automatic strainers For more information, visit www.filtration.eaton.com.
are ideal replacements for both simplex and Or, call 888-309-6648 to speak with an Eaton Oil & Gas Specialist.
duplex manual strainers because of their
reliability and low maintenance.
For more information, call Eaton
at 888-309-6648.
Select 117 at www.HydrocarbonProcessing.com/RS
Select 116 at www.HydrocarbonProcessing.com/RS
 HPINTEGRATION STRATEGIES
DAVE WOLL AND LARRY O’BRIEN, CONTRIBUTING EDITORS

dwoll@arcwb.com

Future of the collaborative process automation system

ARC conceived the Collaborative Process Automation Systems Online version upgrade. Process automation system sup-
(CPAS) in 2002 in response to DCS end users’ requests for us to plier has at least one major version release every year. Unlike a
provide a vision of how process automation systems should evolve. maintenance release, a version release usually requires the system
Clients and others continue to ask ARC where the automation to be taken out of service (thus interrupting production).
market is headed. The DCS came on the scene in 1975 and smart With most of today’s systems, it’s possible to perform a two-
field devices arrived about five years later. Since then, ARC has step version “hot” upgrade without shutting down by loading the
observed incremental innovation, but nothing revolutionary. new version into the redundant side, then committing the new
With this in mind, we recast the request for ARC’s view into the software to the primary side of the system and the process. We
future as an effort to identify innovation we think the market feel that this is unreasonably risky and only marginally practical.
needs to consider in the near future.* We feel a three-step approach where the changes can be validated
is reasonable.
Smart field devices are not smart enough. At present,
50% to 75% of field-device downtime is caused by lack of confi- Business-to-operations integration. Normally, integrat-
dence in the measurements. The current generation of smart field ing two different systems, especially systems as diverse as a business
devices is capable of communicating basic device diagnostics along and manufacturing system, can be complex and difficult. This is
with the measurements. While this has been acceptable in many where Reference Standards can make a major contribution. Their
industries, we would also like to see quantitative measurements primary value is that they contain well-understood and well-doc-
of the relative health of a device and, if there is an issue, how long umented work processes that lead to handing off proper informa-
the device will continue to operate dependably. tion. Several years ago, several senior practitioners on the business
and manufacturing sides recognized the value of integrating Supply
When will distributed control become truly distrib- Chain Management Consortium reference model (SCOR) with
uted? Manufacturing challenges are changing, and automation the manufacturing reference model, ISA-95, and have began to
must change to stay in control. We now deal with a high degree of address the problem.
uncertainty in an increasingly dynamic environment. Manufactur-
ing assets continue to become more costly to purchase. Expansions Application executive of the future. The application
to existing manufacturing assets are expensive, and existing assets executive, a core CPAS function, monitors the health of applica-
almost impossible to replace. While automation can be viewed as tions running in a collaborative and synchronized manner with
a solution to a certain degree, these systems could provide more. other applications, the system itself, and the personnel using the
What have been dubbed distributed control systems in the past system. Today, this is accomplished largely in an environment
are really not all that distributed, since, to a large degree, they still combining proprietary structures, de facto standards and enabling
tie functional requirements to architectural constraints. technologies.
One potential scenario for the future of the application environ-
Future requirements need intelligent agents. Manu- ment is to have this completely based on standards such that no
facturing challenges continue to change, and the current IEC single supplier has to “own” the environment or platform for it to
61131-3 standard will not be able to satisfy the execution seman- function properly. In this future vision, the user would purchase
tics and new requirements for a truly distributed and flexible applications from the supplier that provided the best fit for the
automation system. required functions. It would be loaded into the system and would
The Technical Committee 65 received a new work proposal auto-configure to adapt to the operating system, communications
(NWP) to standardize application of function block software protocols, and whatever hardware requirements are present. HP
modules in distributed industrial-process measurement and con- *Note: Readers can visit www.arcweb.com/res/cpas for more
trol systems. The resulting standard, IEC 61499, will facilitate
information on ARC’s latest research into CPAS.
the manufacturing agents that will shape the next-generation
automation systems.
Dave
Larry WollO’Brien is a vice president
is part of consultingconsulting
of the automation at the ARC Advisory
team at ARC Group. He has
covering the
been associated
process industries, with
anddefining and applyingeditor.
an HP contributing process
He automation
is responsible forfor
over 35 years.
tracking the
Multi-vendor control. The technology exists (or is certainly Mr. Wollfor
is process
currentlyautomation
focused onsystems
assisting major users
market (PASs) and has developing
authored the their
PASprocess
market auto-
stud-
feasible) to make the basic control system a multi-supplier struc- mation
ies system
for ARC strategies.
since 1998. Mr. O’Brien has also authored many other market research,
ture. For example, if the control system LAN were based on strategy and custom research reports on topics including process fieldbus, collaborative
FOUNDATION fieldbus HSE, then just like field instruments Larry O’Brien
partnerships, total automation market
is part of the trends and
automation others. He
consulting has at
team beenARC with ARC since
covering the
January
process 1993, and started
industries. his career with
He is responsible market research
for tracking the marketin the
forfield instrumentation
process automation
can come from different suppliers, the process controllers could
markets.
systems (PASs) and has authored the PAS market studies for ARC since 1998.
also come from different suppliers.
HYDROCARBON PROCESSING NOVEMBER 2010
I 13
Where do You Want to be
on the Performance Curve?

P = People
M= Methodologies
T = Technologies

Your Company + KBC Produces NextGen Performancen

We collaborate with our clients to create unique solutions to their specific
challenges. Some of these challenges may include:
Strategic Challenges Capital Challenges
U Creating Effective Business Strategy/Decisions U Increasing Return on Capital Investment
U Increasing Return on Investments U Rationalising/Optimising Environmental
U Enhancing Returns on Acquisitions/Divestitures Compliance Capital Expenditures
U Reducing Strategic/Capital/Market/Investment Risk U Reducing Capital Risk
U Enhancing Yields
U Creating Effective Response to Crude/ Organisational Challenges
Feedstock/Product Markets U Increasing Organisational Effectiveness
U Improving Financial Performance U Improving Employee Competency/Capability
U Enhancing Employee Support Systems
Operating Challenges U Improving Shift Team Function For more information on how
U Improving Yield
U Increasing Availability Environmental Challenges
KBC can help you achieve
U Reducing Maintenance Costs U Reducing Emissions NextGen Performance,
U Improving Safety Performance U Ensuring Compliance contact us at:
U Implementing/Improving U Reducing/Managing Environmental Liabilities
U Managing Operational Risk U Improving Energy Efficiency
Behaviour-based Reliability
AMERICAS +1 281 293 8200
U Rationalising Compliance Expenditures
U Improving Supply Chain Performance EMEA +44 (0)1932 242424
For 30 years, KBC consultants have provided independent advice and expertise to enable ASIA +65 6735 5488
leading companies in the global energy business and other processing industries manage
risk and achieve dramatic performance improvements. answers@kbcat.com U www.kbcat.com
Select 82 at www.HydrocarbonProcessing.com/RS

ALLAN KERN, GUEST COLUMNIST
Allan.Kern@yahoo.com
Weighing on dual-temperature control
I’d like to weigh in again on the continuing debate regarding
distillation column control.1–4 Attention has now turned to the
practicality of “dual temperature” control,which is the simultaneous
control of temperature on both the overhead and bottoms of a dis-
tillation column, typically with top temperature cascaded to reflux
flow and bottom temperature cascaded to reboiler heat input.
My short answer agrees with Zak Friedman: Beware, all ye who
wander here! Attempting dual-temperature control is most often
undertaken by those oblivious to its perils. Industry is littered
with its wreckage in the form of columns configured for dual-
temperature control, but with one of the controls abandoned—
either the reflux or reboiler is operated in automatic or manual
mode, not cascaded. This is unfortunate from both composition
control and energy conservation points of view.
The lesson from all the wreckage (and from this debate) is that
successful dual-temperature control takes more preparation than
most casual control strategy reconfigurations get. Appropriate
preparation in this case would include a rigorous simulation of the
several variables that affect overall separation because dual-temper-
ature control really means controlling one temperature plus overall
separation (think Δ-temperature between the top and bottom).
For practical purposes, separation is usually considered con-
stant. When Mr. Shinskey or Mr. Friedman points out that a one-
degree change in bottom temperature will result in approximately
a one-degree change in the overhead (Δ-temperature remains the
same), this is another way of saying the same thing. If separation
is constant, it cannot be controlled, and therein is the problem
that many people fail to realize, at least initially. To the extent that
overall separation does vary, it is usually due to uncontrollable fac-
tors, such as feedrate and quality, rather than controllable ones.
That said, if you have the luxury of controlling feed or a feed
analyzer, responsive pressure control, which affects vapor veloc-
ity and tray efficiency, ) excess stages, a reliable simulation, good
operators and economic or equipment limitations that prohibit
doing the split in two columns, then perhaps dual-temperature
control is worth a try. That’s my long answer, which more agrees
with Shinskey, and which can be shortened to: if it must be done,
it might be possible!
A third answer worth mentioning is that the traditional solu-
tion (or compromise) to this dilemma is to put one temperature
control on the top or bottom (whichever purity is the higher prior-
ity or harder to achieve) and operate the other end of the column on
ratio control. This results in good stability, depending on feed and
other variables, and keeps energy consumption in check. The ratio
is usually reboiler heat-to-feed (when top temperature is controlled)
or reflux-to-feed (when bottom temperature is controlled).
The latter option (reflux ratio) has a couple “material balance”
risks. If all the light material is not removed with the distillate, it
can accumulate in the overhead and lead to erratic pressure and
temperature behavior. It’s possible to run the reflux drum empty.
HPI VIEWPOINT
There are no such limitations with the former option (reboiler
ratio), and indeed, it is probably the single most common (success-
ful) column control configuration in the industry. It’s not as good
as the grail of stable dual-temperature control, and is a good mea-
sure better than running the reflux or reboiler at a fixed setpoint,
both in terms of constant separation and energy conservation.
Yet another answer is special cases. On columns that have
side-draws, additional boil-up does not ultimately return as reflux,
so dual-temperature control can work well. This is well known on
main fractionators that often have several temperature control-
lers—top, bottom and side-draws. Occasionally, side-draws are
found in other processes on true distillation columns, as well.
Another special case, although it’s also a trick answer, is dual-
temperature control where the would-be-ratio controller is replaced
with a proportional-only temperature controller with feedforward
signal. This provides for minor adjustment based on temperature
deviation, but without the integral action that leads to wind-up
and instability (although it is still easy to get instability if gain is
set too large). The feedforward signal serves the role of the original
ratio control. In my experience, this does not improve matters
noticeably, but it can please other stakeholders (who don’t “get”
the limitations) to see the controller at least make an effort, and
it makes it easy to switch between top and bottom ratio control,
especially without impacting data historians, documentation and
graphics, since the temperature control tags persist either way.
A final point in the debate has been model-based multivari-
able control (MPC) and whether it changes the game. Arguably,
where dual-temperature control is feasible, model-based control
has a better shot of keeping it in the window, based on its aware-
ness of all the variables affecting separation (not just the relative
gains of the two temperature controllers). But my experience is
that neither MPC nor inferentials (which are essentially tempera-
tures) change the (un)likelihood of success and many late-model
MPCs exhibit similar modern day wreckage in the form of the
reflux or reboiler MVs that are “clamped” or “out-of-service.”
Some things never change! HP
LITERATURE CITED
1 Friedman, Y. Z., “Distillation column dual-temperature control,
Hydrocarbon Processing, March–April, 2010, (letters, August
2010).
2 Kern, A. G., “More on APC designs for minimum maintenance,” Hydrocarbon
Processing, December 2009.
3 Friedman, Y. Z., “APC designs for minimum maintenance,”Hydrocarbon
Processing, June–August, 2009 (and letter by P. R. Latour, October 2009).
4 Greg, F., “Multivariable control of distillation,” ControlGlobal.com, May–July,
2009.
Tim
The Lloyd
authorWright is HP’s European Editor and has been active as a reporter
has 30 years of process control experience and has authored sev-
and conference chair in the European downstream industry since 1997, before
eral papers on inferentials, expert systems, advanced controls and decision support
which he was a feature writer and reporter for the UK broadsheet press and BBC
systems, with emphasis on practical process control effectiveness. Mr. Kern is a
radio. Mr. Wright lives in Sweden and is founder of a local climate and sustainability
professional engineer (inactive), a senior member of ISA, and a 1981 graduate of
initiative.
the University of Wyoming.
HYDROCARBON PROCESSING NOVEMBER 2010
I 15

BILLY THINNES, NEWS EDITOR
BT@HydrocarbonProcessing.com
LCFS will adversely affect
oil-sands crude refiners
Purvin & Gertz, Inc., has released
a report on the effect low-carbon-fuel
standards (LCFS) will have on oil sands.
The report notes that LCFS programs
are being implemented in California,
Oregon and British Columbia. They
are under consideration in many other
states and provinces and are becoming
regional in nature. LCFS programs dif-
fer by jurisdiction, but have in common
mandated reductions in the carbon inten-
sity of transportation fuels. By targeting
petroleum-derived gasoline and diesel and
promoting low-carbon alternative energy
forms such as electricity, hydrogen, natu-
ral gas and next generation biofuels, LCFS
programs are intended to reduce overall
greenhouse gas (GHG) emissions on a
“well-to-wheels” basis.
“In effect, LCFS programs contribute to
an ‘off-oil’ strategy,” said Tom Wise, who
directed the study.
The energy needed to produce oil
sands crudes is higher than for most con-
ventional crudes, so the resulting carbon
intensities of refinery-produced gasoline
and diesel from oil sands are also higher.
Tom Wise points out it is a mistake to
paint all oil sands crudes with the same
brush because there are different oil sands
crudes, such as synthetic crude oil and
bitumen blended with various diluents,
and each has a different pathway and car-
bon intensity.
“Contrary to widely held perceptions,
our study concludes that some oil sands
diluted bitumen does not have high carbon
intensity under the California regulations
and should not carry an LCFS penalty,”
Mr. Wise said.
The Purvin & Gertz study estimates
the well-to-wheels carbon intensities of
refinery-produced gasoline and diesel from
various oil sands and conventional crude
oils. The study estimates the impact on
consumer product prices, refinery mar-
gins and oil sands crude prices, for a range
of LCFS carbon costs. In market regions
that implement LCFS programs, consumer
product prices will increase and refinery
margins will fall.
“Some of the oil-sands crudes would
require price discounts to compete with
conventional crudes due to a reduced incen-
tive to refine or upgrade heavy crudes,” said
Mr. Wise.
Further, reduced crude runs in market
regions with LCFS programs could result
in refinery closures and displace oil sands
crudes to other markets. For instance,
LCFS programs in the US Midwest would
cause leakage of oil sands crudes to the US
Gulf Coast or Asia-Pacific.
US demand for lubricants
to reach 2.25 billion
gallons in 2014
US demand for lubricants is forecast
to expand 1.3% annually to 2.25 billion
gallons in 2014, valued at $22 billion.
This represents a significant improvement
over the performance of the 2004–2009
period, when lubricant demand declined
5% annually. A turnaround in motor
vehicle production, along with an accel-
eration in the number of automobiles in
use, will support demand for automotive
lubricants. Additionally, increased manu-
facturing output will drive demand for
industrial lubricants. However, total lubri-
cant consumption is not expected to reach
pre-recession levels. This will largely be
due to the greater use of longer-lasting,
higher-performing synthetic lubricants
that extend drain intervals, therefore
reducing overall lubricant requirements
in volume terms. Average price increases
will continue to be significant due to
expected growth in crude oil prices and a
shift in product mix toward higher-value
lubricants. These and other trends are pre-
sented in a new study from The Freedonia
Group, Inc.
Engine oils accounted for more than half
of total US lubricant demand in volume
terms during 2009. A significant rebound
in motor vehicle output following the dou-
ble-digit annual declines of the 2004–2009
period will propel engine oil demand in
the factory-fill segment. However, this rep-
resents only a small fraction of engine oil
demand, and the overall outlook for these
products will be restricted by lengthening
oil change intervals and the use of high
HPIMPACT
performance synthetic lubricants. As such,
aftermarket demand will decline, with the
“do-it-yourself” segment continuing to lose
out to “do-it-for-me” services, a trend that
stalled in 2008 and 2009 as drivers sought
out more economical alternatives for their
vehicle service needs during difficult eco-
nomic times.
Process oils—including white oils,
rubber oils, electrical oils, ink oils, agri-
cultural spray oils and defoamer oils—rep-
resent another leading lubricant category.
Demand for these products is forecast to
advance at the most rapid pace, promoted
by rebounding manufacturing activity fol-
lowing the real (inflation-adjusted) declines
of the 2004–2009 period. In particular, an
improved outlook for food and beverages,
chemicals, plastics and rubber will offer
good growth opportunities. However, pro-
cess oils will continue to encounter chal-
lenges brought about by changing environ-
mental and regulatory standards.
Linde supplying
hydrogen technology
to US BMW plant
The Linde Group will provide the
BMW Manufacturing Co. plant in Spar-
tanburg, South Carolina, with a hydrogen
fueling system for its material-handling
fleet. An according agreement was signed
in mid-August 2010. More than 85 pieces
of material-handling equipment are hav-
ing their lead acid batteries replaced with
hydrogen fuel cells. The trucks deliver
process parts to assembly areas throughout
the plant. After the conversion, this part of
BMW’s internal logistics will be completely
emission-free.
“This is one of the largest hydrogen
applications of its kind,” said Dr. Andreas
Opfermann, head of innovation manage-
ment of The Linde Group. “We are proud
to work together with BMW, supplying
both a high-efficiency fueling system and
hydrogen with almost no carbon footprint.
This project clearly shows the potential that
hydrogen offers for internal logistics.”
Refueling at the six indoor dispenser
stations will be facilitated by Linde’s ionic
compressor fueling system, combining
efficiency and high throughput with low-
HYDROCARBON PROCESSING NOVEMBER 2010
I 17
HPIMPACT
maintenance costs and low noise compres-
sion. The hydrogen supplied to BMW, a
byproduct from a chemical plant, is puri-
fied, compressed and liquefied by Linde
using electricity produced from renewable
hydropower.
“We have a clear vision and we are deter-
mined to reach our goal of using renewable
energy as much as possible throughout the
plant site,” said Josef Kerscher, president of
BMW Manufacturing. “We realize this is
Don’t let Facility Siting negatively
impact your worker efficiency or plant safety.
The #1 supplier of protective building systems,
and the leader in customer VALUE:
an ambitious goal and the hydrogen fuel
cell project puts us another step toward ful-
filling our vision.”
BMW is able to boost productivity
in two ways: Refueling these trucks with
hydrogen takes an operator less than three
minutes, compared with about 20 minutes
to change out a battery. These units, when
full, run on 2 kilogram fuel tanks and oper-
ate for 8–10 hours. Also, fuel cells do not
lose power over time, as lead-acid batteries
do toward the end of a shift. BMW is also
able to reduce its total electricity demand,
since no battery recharging is required, and
eliminate the environmental disposal costs
for lead acid batteries.
Oil and gas reserves
increased 3 percent
in 2009
The worldwide upstream investment of
224 oil and gas companies decreased 23%
to $378 billion in 2009, according to a
report released by IHS Herold. Although
development spending fell nearly 20%, the
first decline in a decade, total hydrocarbon
reserves increased 3% as both oil and gas
reserves grew for the first time since 2005.
Production also increased 1%, driven by a
2.2% increase in natural gas output.
“We were very surprised at the strength
of reserve additions given the weak eco-
nomic conditions and tightness in credit
markets during 2009,” said Nicholas Cac-
chione, director of IHS Herold and author
of the report. “As an industry, we spent
fewer dollars, but they went further in
terms of purchasing power.”
Oil reserves reversed a two-year decline,
rising 3% to 164 billion barrels. The main
driver was 8.6 billion barrels in positive
reserve additions, but extensions and dis-
coveries in the Canadian oil sands and
South and Central America also added
a record 7.9 billion barrels. Natural gas
reserves climbed 3.7% despite a record 11.4
trillion cf in negative reserve revisions, as
development of unconventional plays in

V ariety of standard & custom building options for permanent applications North America and liquefied natural gas
(LNG) resources in Asia accelerated.

A vailability of quick response buildings for lease applications The decline in capital spending was led
by a 40% reduction by E&P companies,

L ogistically friendly utility units that do not require the use of a crane while the integrated oil companies cut
investment by just 9%. Exploration spend-

U nique fastening system that significantly reduces installation time

ing was most resilient, dropping just 12%
to $62.7 billion. In contrast, unproved

E xtensive manufacturing capacity (1.6m sq.ft., strategically located in 3 countries)

acquisition costs were down 71%, and
a 2% dip in proved acquisition outlays
would have fallen 50% were it not for the
$20 billion Suncor/Petro-Canada merger.
“With the recession and ongoing uncer-
tainty in the market last year, companies
put acreage acquisition on hold and seemed
to focus on their in-house development
purchase, lease, and lease to purchase opportunities,” said Mr. Cacchione. “This
ISO 9001 Registered decision, I think, reflected their desires
to monetize known holdings that can be
brought into production much more rap-
BlastResistantBuildings.com idly than something with a less certain pay-
out several years down the road.”
Select 152 at www.HydrocarbonProcessing.com/RS
18
Make a smart turn and
take days off your turnaround.

BJ Services’ new suite of SmartTurn™ services is the shortest route

to improved efficiency and time saving during plant turnarounds.
BJ Services’ SmartTurn™ engineered approach reduces the time and effort required to cool down and
clean equipment, extending time available for repairs, inspections and upgrades. Using industrial gases,
BJ technicians cool reactors and furnaces in hours, not days. Our cryogenic industrial cooldowns have
been proven in both sulfur recovery and ethylene production units.
BJ Services’ applied chemical procedures remove contaminants and volatile gases from piping, heat
exchangers and towers to reduce, or eliminate, the need for mechanical cleaning of crude, asphalt and
FCC heat exchangers. SmartTurn chemistry also removes hydrocarbon contamination, hydrogen sulfide
and pyrophoric contaminants.
BJ Services’ years of chemical and engineering experience gets treated equipment ready for entry and
hot work quickly. So, take the right turn with BJ’s SmartTurn approach to plant turnarounds.

Real world. World class. Worldwide.

www.bjservices.com

Select 69 at www.HydrocarbonProcessing.com/RS
HPIMPACT
Lower capital spending and higher
reserves resulted in a near 50% decrease in
reserve replacement costs (to $11.41/boe)
and lowered finding and development costs
to $12.23/boe. Strong natural gas reserve
additions led reserve replacement rates to
the highest levels in five years.
Despite the strong performance metrics,
upstream profits plunged 47% as a 13%
decline in pre-tax expenses did not offset a
30% reduction in revenues. The integrated
Pflj\\jk\\c%
N\j\\jX]\kp%
?lek\i9l`c[`e^j`jk_\nfic[c\X[\i`e
?lek\i9l`c[`e^j`jk_\nfic[c\X[\i`e
k_\gif[lZk`fef]hlXc`kp$\e^`e\\i\[#
k_\ gif[lZk`fe f] hlXc`kp$\e^`e\\i\[ 
YcXjk$i\j`jkXekdf[lcXiYl`c[`e^j%
Gligfj\Yl`ck]ifdk_\^ifle[lg#\m\ip
?lek\iYl`c[`e^d\\kjXe[\oZ\\[j`ek\ej\
jX]\kpXe[YcXjkjkXe[Xi[jkf\ejli\k_\
gifk\Zk`fec\m\cjpfl`ej`jkfe]fig\ijfee\c#
Zi`k`ZXc\hl`gd\ekXe[gif[lZk`m`kp%
›:ljkfd;\j`^e
›J_fik$fiCfe^$K\idC\Xj`e^
›G\idXe\ekK\dgfiXip8ggc`ZXk`fej
›Dlck`gc\9cXjkFm\igi\jjli\;liXk`fe;\j`^ej
›Cfn#D\[`ldXe[?`^_I\jgfej\;\j`^ej
›Gfj`k`m\Gi\jjli`qXk`fe
›:cXjj@;`m`j`fe)<c\Zki`ZXc
›>Xj=`i\;\k\Zk`fe
›JX]\?Xm\e:XgXYc\
›KiXejgfikXYc\n`k_Hl`ZbJ\klg
›8G@IG.,)&.,*:fdgc`Xek
›9cXjkK\jk\[
J\\jX]\kp]ifdflij`[\ Ç ?lek\i9l`c[`e^j
b\\gjpflgifk\Zk\[]ifdk_\flkj`[\`e%
_lek\iYl`c[`e^j%Zfd
Select 153 at www.HydrocarbonProcessing.com/RS
20
oil companies accounted for 85% of the
universe’s profit with the E&P companies
accounting for the balance. Reserve write-
downs slashed net income for the large E&P
companies and drove the mid-sized and
small E&P companies to a loss. However,
the industry generated free cash flow due to
the steep decline in capital investment.
The IHS Herold report found that
dividends rose modestly to another record
level, which it noted “is remarkable” given
)/(%+,)%0/''
the turmoil in the financial markets and
the generally miserable results in the indus-
try’s downstream operations. Dividends
exceeded $100 billion, but common share
repurchases were 23% lower, falling for the
first time since 2004. Capital constraints
brought about by reduced revenue and ris-
ing costs have almost completely eliminated
share buybacks as a viable use of funds.
Key regional findings of the report
include:
• Strong drill-bit additions aided
improving results for reserve replacement
costs and rates in the US. Unit profitabil-
ity declined for the fourth consecutive
year. Mineable bitumen reserve additions
in Canada offset weak natural gas reserve
additions. Profits were down sharply in
Canada as well.
• Oil and gas reserves in Europe con-
tinued to decline as companies redirect
cash flows to other regions. The reserve
replacement rate reached a five-year high
through improving reserve additions,
but the region was still below full reserve
replacement figures.
• Capital spending in the Africa and
Middle East regions was down 14%, which
was much less than the worldwide average.
This drop in spending is due to regional
dominance by the integrated oil compa-
nies, that tend to spend through the com-
modity cycles.
• Asia-Pacific reserves gained 3% as nat-
ural gas extensions and discoveries surged.
Reserve replacement rates in the region
were well above full replacement levels.
• Capital spending in South and Cen-
tral America increased since regional play-
ers have strong development portfolios to
exploit. Total reserves in the region increased
3%, the first gain in several years.
• An uptick in proved acquisition spend-
ing limited total capital spending decline to
17% in the Russia/Caspian region, while
drill-bit spending outlays fell 22%. Pro-
duction in the region increased 18%, with
strong results from both oil and gas output.
IHS Herold anticipates a modest global
rebound in upstream capital spending in
2010.
“In North America, E&P investment
increased 30% in the first half of 2010,
which was higher than expected,” Mr.
Cacchione said. “We think this should
drive a global investment increase of more
than 20% for the year. Outside of North
America, where spending declines were less
severe, we foresee upstream investment ris-
ing about 10%.” HP
Sharper Detection of
Off-Angle Cracks Leak
Repairs

Team Phased Array Scanner Improves Inspection Results Field Heat

Treating

Field
Machining

NDE/NDT
INSPECTION

Hot Taps /
Line Stops

Technical
Bolting

Valve
Repair

Valve
Insertion

T he world leader in NDE/NDT Inspection, Team now

brings you the Phased Array Scanner. Lower costs,
higher quality, greater speed … you get all the advantages
Emissions
Control

of phased array in a highly advanced system.

Pipe Repair
Services
Driven by high-speed electronics and real-time imaging,
Team’s Phased Array Scanner reveals the precise location
and size of off-angle cracks, f laws and defects. Inspection
time and costs are reduced as Team technicians produce
fast, detailed cross-sectional images of welds and other
internal features. Flaws or defects are measured accurately
and marked for repair immediately following examination.
 
From small boiler tubes to massive turbines and vessels,
Team’s Phased Array Scanner saves you time and money by
delivering fast, accurate results. Call +1-800-662-8326 or
visit www.teamindustrialservices.com/phasedarray.
Select 73 at www.HydrocarbonProcessing.com/RS
Select 83 at www.HydrocarbonProcessing.com/RS

SELECTED BY HYDROCARBON PROCESSING EDITORS
editorial@gulfpub.com
New FCC catalyst announced
Grace Davison has released a new fluid
catalytic cracking catalyst technology called
Alcyon. This catalyst reflects Grace’s con-
tinuing commitment to delivering a broad
portfolio of FCC catalysts and additives.
Alcyon is designed for applications that
require maximum activity and controlling
coke make.
Alcyon catalyst contains a proprietary
zeolite modification that delivers the high-
est activity of any catalytic technology
with a low surface area per kinetic conver-
sion. An added benefit of the new zeolite
is superior activity retention that has been
verified in field performance. At constant
surface area, Alcyon is more active than
a traditional catalyst; thus it requires less
catalyst-to-oil (C/O) ratios to achieve
desired conversion levels and products.
This new FCC catalyst uses improved
access to active sites with increased crack-
ing rates per active site.
It is the latest of 11 new refining prod-
ucts introduced since 2008 by our research
and development group,” commented
Shawn Abrams, vice president and general
manager of Grace Davison Refining Tech-
nologies. “Like all our FCC catalysts and
additives, Alcyon catalyst’s revolutionary
technology and performance are supported
by our strong technical service and flexible
manufacturing system.”
Alcyon is particularly well suited for
refiners looking to reoptimize unit opera-
tion to maximize profitability, yet remain
within operating limits. For any given coke
yield, Alcyon catalyst achieves higher con-
version, maximizing total FCC barrels to
the refinery gasoline pool.
Select 1 at www.HydrocarbonProcessing.com/RS
Collaboration improves
emergency valve performance
ABB, the leading power and automa-
tion technology group, announced that it
will collaborate with Dresser Masoneilan,
a global leader in process control valves,
on an integrated process to monitor, test
and manage emergency shutdown valves
(ESDVs) during all operational condi-
tions, from normal plant operations to
abnormal situations. These valves are cru-
cial process elements for the oil, gas and
HPINNOVATIONS
petrochemical industries, as well as for
many other industrial processes.
“Our collaboration with industry lead-
ers like Dresser Masoneilan helps us to
offer our mutual customers best-in-class
safety solutions that will protect the integ-
rity of their processes and the surrounding
community,” said Luis Duran, Americas
business development manager for Safety
Systems, ABB.
The combined solution leverages the
capabilities of ABB’s 800xA high-integ-
rity safety instrumented system (SIS)
and Masoneilan’s SVI II emergency shut-
down device (ESD) and PST controller to
improve overall plant safety and increase
the availability of ESDVs for optimal
response of the isolation valve in emer-
gency situations. This integration also sim-
plifies safety compliance by automatically
recording partial stroke test results and
emergency shutdown events, saving time
and money while increasing efficiency.
“By taking advantage of System 800xA’s
unique integration capabilities and open
standards, the user has immediate access
to the health diagnostics and status of the
emergency shutdown valve. This access
also provides proactive management of
this critical device, for instance, enabling
remote triggering of partial stroke tests,
to ensure that it is ready to perform when
needed,” said Kristian Olsson, manager of
ABB’s Safety Center of Excellence. “This
immediate readiness is vital to the protec-
tion of the process, the environment and
the surrounding community in the event
of an abnormal situation.”
As an integrated object within System
800xA, Masoneilan’s SVI II ESD device
can be configured to perform scheduled
partial valve stroke tests while remotely
monitoring and maintaining the emer-
gency shutdown valves during normal
plant operations. This minimizes the need
for outages and downtime to evaluate the
health and readiness of these critical pro-
cess elements. This also provides easy-to-
understand alerts and recommendations
regarding valve status, as well as required
partial stroke test and emergency shutdown
signatures and documentation.
“While open standards offer great ben-
efits for end users, it is the collaboration
between automation vendors that provides
for an ‘out-of-the-box’ solution capable
of generating instant results,” said Sandro
Esposito, global marketing manager of dig-
ital products for Dresser Masoneilan. “The
SVI II ESD provides an excellent return on
investment with its combined shutdown
function, partial stroke test function and
shutdown event ‘blackbox’ into a single
SIL3-certified device.”
The SVI II ESD is the latest technology
in emergency shutdown valve automation
and in-service valve partial stroke testing.
The SVI II ESD is the only SIL3-certi-
fied ESD certified at 4mA with stainless
steel housing. The device can be imple-
mented using a 4/20mA signal (analog
safety demand), 0-24Vdc (discrete safety
demand) or a combination of both. Stan-
dard on the device are an LCD display
and explosion-proof external pushbut-
tons. This design architecture offers a
sophisticated platform while being Type
A (simplex device) compliant. System
800xA high integrity is ABB’s next gen-
eration safety system. This SIL 3-rated SIS
provides the highest level of integration
of safety and control on the market and
a unique embedded diverse technology
architecture that provides superior protec-
tion of the process, plant, personnel and
the environment while it optimizes overall
process efficiency.
Select 2 at www.HydrocarbonProcessing.com/RS
Gasoline benzene reduction
alkylation catalytic technology
Today, more stringent regulation is
challenging refiners to meet the tighten-
ing gasoline specifications for benzene at
As HP editors, we hear about new
products, patents, software, pro-
cesses, services, etc., that are true
industry innovations—a cut above the
typical product offerings. This sec-
tion enables us to highlight these
significant developments. For more
information from these companies,
please go to our website at www.
HydrocarbonProcessing.com/rs and
select the reader service number.
HYDROCARBON PROCESSING NOVEMBER 2010
I 23
Tower Power

The new Baldor RPM AC® Cooling Tower Direct systems, the RPM AC Cooling Tower motor and • Energy Efficient
Drive Motor sets a new standard in cooling VS1CTD Cooling Tower drive offers improved
tower motors. Designed exclusively for cooling reliability while greatly reducing maintenance • Unmatched Quality
tower applications, the RPM AC motor mounts costs and energy consumption.
directly to the cooling tower fan, eliminating
• Superior Reliability
high-maintenance gearboxes, drive shafts and Check out our website for case studies,
couplings. Combined with Baldor’s VS1CTD installation photos and additional information
proprietary adjustable frequency drive, this slow about the revolutionary new RPM AC Cooling
• Low Maintenance
speed, high torque combination operates at Tower technology from Baldor. We are truly
variable speeds to maximize system efficiency changing the future of cooling tower technology! • Quiet Operation
and minimize noise. Perfect for new installations
or for retrofitting older, less efficient cooling tower baldor.com/CoolingTower 479-646-4711 • Made in the USA

©2010 Baldor Electric Company

Select 80 at www.HydrocarbonProcessing.com/RS

the lowest cost and without significant
octane loss. For instance, the US Environ-
mental Protection Agency’s most recent
clean fuels regulations (Mobile Source
Air Toxics II) require refineries to reduce
benzene to less than 0.62 vol% (on an
annual basis) in gasoline by 2011, from
its current level of 1.0 vol%.
The limits apply to both reformu-
lated and conventional gasoline. Sources
of benzene in the gasoline pool vary for
each refinery; the predominant source for
most is reformate. Benzene reduction in
reformate can be achieved by either the
removal of benzene precursors before they
are converted to benzene in the reformer,
or by post-removal of benzene in the
reformate product by chemical conver-
sion of benzene or by removal with frac-
tionation.
EMRE developed BenzOUT, a refor-
mate alkylation technology, to convert
benzene into high-octane alkylaromatic
compounds (such as isopropylbenzene)
for gasoline blending by reacting a ben-
zene-rich stream with light olefins such
as propylene in low-value olefin streams.
This patented technology avoids the
octane loss and hydrogen consumption
associated with the alternative option of
benzene saturation.
BenzOUT is a refining process based on
ExxonMobil’s ethylbenzene and cumene
technologies, which have been widely
used in the chemical industry with over
80 worldwide commercial applications.
In addition, EMRE’s BenzOUT tech-
nology was commercially demonstrated
at a North American refinery. BenzOUT
reduces benzene by reacting a benzene
concentrate stream with a light olefin-
containing stream such as LPG over a
proprietary catalyst. Typically, a benzene
concentrate ranging from 10% to 50% is
processed.
Key features of the process include:
Fixed-bed catalyst technology—This
process uses a simple fixed bed reactor.
In revamp projects it is possible to ret-
rofit existing polygas tubular/chamber
reactors or spare reformer reactors for this
application.
Catalyst—This process utilizes a pro-
prietary solid acid catalyst with long cycle
lengths. It possesses high catalyst activity
and allows long catalyst cycle life. In addi-
tion, the catalyst is completely regenerable
ex-situ to further extend catalyst life.
Feed requirement—This process requires
conventional feed pretreatment for the
HPINNOVATIONS
olefin streams to remove potential contami- specific volume swell will be dependent
nants such as sulfur and nitrogen species. on the feed composition and level of ben-
Some process advantages include: zene conversion.
Benzene reduction—Reduction of ben- Octane gain—2–5 numbers of (R+M)/2
zene content in the gasoline pool to meet increase is typical. The specific octane gain
new benzene regulation. High benzene depends on the feed composition.
conversion can be achieved. Reformer flexibility for increased hydrogen
Gasoline volume swell—Upgrading of production.
light olefins and benzene into high-octane Select 3 at www.HydrocarbonProcessing.com/RS
gasoline blend stock also results in a vol-
ume swell of the gasoline volume. The
Simulation to Business. Knowledge to Profit.
Do all safety valves open
simultaneously?
Dynamic Simulation following the API
guideline can help you to reduce Capex
during flare system debottlenecking -
accounting for the real transient beha-
viour of your plant during emergencies
or scheduled blow-downs.
SIMULTANEITY EFFECTS OF A NUMBER OF PROCESS UNITS
Sum (t)
FCC
Flare load, t/h
Coker 1
CDU 2
Time, min ?khf3@kn[^k^m'Ze'3 Are there alternatives to an expensive
overhaul of a bottlenecked flare system (PTQ Q1 2010)
Experienced Inprocess Engineers support
you with accurate analysis of dynamic
emergency and blow-down behaviour for:
Dynamic flare load calculations
Dynamic blow-down / depressuring calculations
Flare System debottlenecking
Recalculation of safety systems / HIPPS
Simulation Simulation Software Operator
Training Services & Extensions & Training
Programs Applications Programming Programs
inprocess Technology and Consulting Group, S.L.
Gran Via de Carles III, 86 - Torre Est, 9è-1a 08028 Barcelona. Spain M^e3+34 933 308 205 ?Zq3 +34 933 308 206
^fZbe3 info@inprocessgroup.com ammi3((bgikh\^ll`khni'\hf
Select 154 at www.HydrocarbonProcessing.com/RS
25
 Are You Prepared for CFATS?
Let ADT Advanced Integration Help.

We have been a trusted source in the petrochemical security business long before 9/11,
MTSA and CFATS: Chemical Facility Anti-Terrorism Standards regulations.

ADT has full SAFETY Act certification and designation for Electronic Security Services from the
U.S. Department of Homeland Security. Our Advanced Integration Division’s Petro-Chem & Energy
Solutions Team can provide:
• A total security management plan: perimeter detection including fiber-based fence detection,
thermal video analytics, remote video monitoring, maritime radar, K-Rated vehicle barriers & fencing
and access control
• System consultation, project management and coordination, system installation and commissioning,
general construction, system training, and maintenance and service

We help protect 13 of the top 20 global petrochemical and energy companies. To learn more, call
1-888-228-0227 or download your FREE copy of Ten Tips for Completing a CFATS Site
Security Plan at www.ADT.com/whitepaper
Select 54 at www.HydrocarbonProcessing.com/RS

PERIMETER DETECTION VIDEO ANALYTICS CENTRAL COMMAND & RESPONSE K-RATED & HYDRAULIC GATE ENTRY

ADT state license numbers are available for review at www.ADT.com or by contacting 1-800-ADT-ASAP.® ©2010 ADT Security Services, Inc. All Rights Reserved. ADT and the ADT logo are registered trademarks of ADT Services, AG,
and are used under license. ADT Advanced Integration is a division of ADT Security Services, Inc.
 HPIN CONSTRUCTION
BILLY THINNES, NEWS EDITOR

BT@HydrocarbonProcessing.com

North America BlueFire Renewables, Inc., has final- South America

Flint Energy Services Ltd. was awarded ized and signed an engineering, procure- GE Oil and Gas has been awarded a
a contract extension for work on Suncor ment and construction (EPC) contract $32 million contract by PlusPetrol Peru
Energy’s Firebag SAGD projects near Fort for its planned cellulosic ethanol facility Corporation SA to supply compression
McMurray, Alberta, Canada, for site-wide in Fulton, Mississippi. The facility will be and reinjection upgrade equipment to boost
construction and commissioning services. engineered and built by Wanzek Con- the output of the Malvinas natural gas liq-
The contract extension of approximately struction, Inc., a wholly owned subsidiary uid (NGL) processing plant, located 400
$78 million will increase the company’s of MasTec, Inc., for a fixed price of $296 km from Lima at the southern section of
total contract scope to just over $110 mil- million which includes an approximately the Peruvian Amazon rainforest.
lion. Work on the contract will be com- $100 million biomass power plant as part Under the contract, GE will upgrade two
pleted by Flint’s Facility Infrastructure of the facility. existing GE gas turbine modules with GE
Division and will employ over 700 workers The contract is negotiated in a manner technology to reduce emissions and other
through early 2011. to be appealing for non-recourse project technology to reduce outages and extend
bank financing and, more importantly, the life cycle of the equipment. In addition,
KP Engineering, LP, recently com- serves as the final key project contract agree- three existing modules will be upgraded in a
pleted a 13,000-bpd FCC gasoline ment to move forward with both the DOE different fashion. The equipment was man-
hydrotreater for Wynnewood Refining and USDA loan guarantee programs. ufactured in Florence, Italy, and shipped to
Co. in Wynnewood, Oklahoma. The proj- The project will allow BlueFire to uti- Peru in March 2010. With the use of trucks
ect, which was executed through an engi- lize green and wood wastes available in the prohibited in the environmentally sensitive
neering, procurement and construction region as feedstock for the ethanol plant region, GE overcame the logistical challenge
management (EPCM) contract, achieved that is designed to produce approximately of “on-time” delivery by taking advantage of
mechanical completion in September. 19-million gpy of ethanol. seasonal swelling of the Amazon and Uru-
The hydrodesulfurization unit utilizes the bamba rivers to transport the two 60-ton
Prime-G+ process, licensed from Axens Technip has two lump-sum contracts modules to the Malvinas plant.
North America, Inc. The unit is designed with Valero Refining Co. and Diamond
to produce FCC gasoline with a sulfur con- Shamrock Refining Co. for two hydrogen Foster Wheeler AG’s Global Engineer-
tent of less than 50 ppm and is a key com- plants at the company’s refineries in Mem- ing and Construction Group has been
ponent of Wynnewood’s overall strategy for phis, Tennessee, and McKee, Texas. The awarded a contract by YPF SA for a new
TREND ANALYSIS FORECASTING
clean transportation fuels. Unit startup is two 30 million standard cubic feet per day delayed coking unit to be built at YPF’s
Hydrocarbon Processing maintains an
scheduled for the fourth quarter of 2010. hydrogen plants will produce high purity
extensive database of historical HPI proj-
Complejo Industrial La Plata in Argentina.
hydrogen and export
ect information. steam.
Current Theactivity
project plants Foster Wheeler’s scope of work includes
Gulf Coast Fractionators, a partnership will use a high
is published efficiency
three times a top-fired
year in thesteam
HPI
among ConocoPhillips, Devon Energy Construction
methane Boxscore.
reforming When
process and autilize
projectthe
Corp. and Targa Resources Partners LP, is completed, it is removed from current
latest nitrogen oxide reduction technology TREND ANALYSIS FORECASTING
listings and retained in a database. The
announced plans to expand the capacity of thereby ensuring minimum emissions. Hydrocarbon Processing maintains an
database is a 35-year compilation of proj-
extensive database of historical HPI proj-
its natural gas liquids fractionation facility Thebycontracts
ects cover basic
type, operating engineering,
company, licen-
ect information. The Boxscore Database is a
located in Mont Belvieu, Texas. The maxi- sor, engineering/constructor,
project management, detaillocation, etc.
engineering, 35-year compilation of projects by type, oper-
Many companies use the historical data for
mum gross fractionation capacity of the fabrication, supply and installation, pre-
trending or sales forecasting.
ating company, licensor, engineering/construc-
facility will be expanded by approximately commissioning and startup assistance. tor, location, etc. Many companies use the his-
The historical information is available in torical data for trending or sales forecasting.
42% (43,000 bpd) to 145,000 bpd. Technip’s operating
comma-delimited center
or Excel in Claremont,
® and can be cus- The historical information is available in
ConocoPhillips, as the operator, will California,
tom sorted will execute
to suit theseThe
your needs. contracts.
cost of comma-delimited or Excel® and can be cus-
manage the expansion project and exist- the sort is
Technip depends on the
partnered size Performance
with and complex- tom sorted to suit your needs. The cost of the
ity of the sort you request and whether a
ing operations are not expected to be dis- Contractors Inc. for the installation
customized program must be written. You
sort depends on the size and complexity of
the sort you request and whether a custom-
rupted during the construction phase. The ofcan
both hydrogen
focus on a narrow plants. Thesuch
request project
as theis ized program must be written. You can focus
expansion is expected to be operational scheduled
history oftoa be completed
particular typeinofthe first half
project or on a narrow request such as the history of a
during the second quarter of 2012, subject ofyou can obtain the entire 35-year Boxscore
2012. particular type of project or you can obtain
database, or portions thereof.
to regulatory approvals. The total capital the entire 35-year Boxscore database, or por-
expenditures of approximately $75 million Simply send Group,
Willbros a clear description of the
Inc., has been data
selected tions thereof.
you need and you will receive a prompt Simply send a clear description of the data
are expected to be significantly lower than Diamond
bycost Shamrock
quotation. Contact: Refining Co. to you need and you will receive a prompt cost
a greenfield fractionation facility since the provide engineering, procurement and con-
Lee Nichols quotation. Contact:
new capacity will be integrated with the struction (EPC) P. O. Box new
of the 2608flare gas recov- Lee Nichols
existing fractionation capacity, utilities, ery facilities Valero
at theTexas,
Houston, McKee refinery
77252-2608 P.O. Box 2608, Houston, Texas, 77252-2608
infrastructure and footprint already at Fax: 713-525-4626
located in Sunray, Texas. The project is val- Fax: 713-525-4626
Mont Belvieu. ued ate-mail: Lee.Nichols@gulfpub.com.
approximately $14 million. e-mail: Lee.Nichols@gulfpub.com

HYDROCARBON PROCESSING NOVEMBER 2010

I 27
HPIN CONSTRUCTION
detailed engineering, procurement services
and support to construction and plant
startup. The unit will use Foster Wheeler’s
delayed coking technology. The planned
new facility will have a capacity of 28,000
bpsd, and will produce anode coke to be
used in the aluminum industry.
Europe
Avantium has started with the construc-
tion of a pilot plant at the Chemelot site in
Geleen, The Netherlands. The pilot plant
will produce building blocks for making
green materials and fuels. Avantium has
developed a proprietary catalytic process to
convert carbohydrates into furanic build-
ing blocks under the brand name YXY.
The pilot plant is expected to become
operational in the first quarter of 2011.
It will demonstrate the process developed
in Avantium’s labs on a larger scale. Fur-
thermore, it will produce several tons of
furanic building blocks per year to support
product development. Avantium is col-
laborating with industrial partners such as
NatureWorks (a subsidiary of Cargill) and
Teijin Aramid to develop novel materials
on the basis of its furanic building blocks.
Avantium collaborates with DAF Trucks
(a Paccar company) on the development of
furanic building block fuels.
The pilot plant is partly funded by a €1
million grant from the Dutch Ministry of
Agriculture.

Middle East
Siirtec Nigi S.p.A. has been awarded
two contracts by Saipem S.p.A. to design
and supply eight packages, each including
a burner, reaction furnace, steam drum,
waste heat boiler, and 20 sulfur condensers.
This equipment represents the core of the
sulfur Claus unit. It will be installed at the
Shah gas field in the UAE. The field units
will process around 1 billion scfd of sour
gas to produce around 540 million scfd of
gas suitable for consumption.
Siirtec Nigi’s scope of work includes the
supply of know-how, design and supply of
the equipment together with the relevant
process guarantees. In addition, Siirtec Nigi
will provide site services and specialists for
commissioning and startup.

Asia-Pacific
Toyo Engineering Corp. has a con-
tract for an ethanolamine (ETA) produc-
tion plant to be constructed in Jiaxing,
Zhejiang Province, China, by Honam
Petrochemical Corp. (HPC). The plant is
scheduled for completion in the first quar-
ter of 2012.
The ETA project is the third contract
that Toyo has received from HPC, follow-
ing a polypropylene plant project awarded
in 1977 and an ethylene plant project in
1989. The past successful projects and sup-
This bench top analyzer tops all others in its price range for ports from Toyo Group, including Toyo-
features and performance. It’s equipped with an intuitive user Korea, have led to the award. The plant
interface, full-color touch screen and on-board Windows XP will produce ETA. As a result of bidding,
computer. Ethernet electronics that permit remote access for Toyo-China has won the engineering, pro-
calibration, diagnostics or service support. Plus, the Phoenix II curement and construction (EPC) contract
has a large sample compartment that accommodates spinners after successfully competing against other
and special holders yet requires little or no sample preparation. contractors in China and Korea.
It all adds up to the lowest cost of ownership, backed by
AMETEK’s reputation for reliability and world class customer China Petroleum and Chemical Corp.
support. Visit: ametekpi.com plans to build a liquefied natural gas (LNG)
receiving terminal in Maoming, Guangdong
Province, China. The construction project
will include an oil and gas wharf, an LNG
receiving terminal and 500,000-cubic-
meter LNG storage tanks. HP
Select 155 at www.HydrocarbonProcessing.com/RS
28
 HPI CONSTRUCTION BOXSCORE UPDATE
Company City Plant Site Project Capacity Unit Cost Status Yr Cmpl Licensor Engineering Constructor
AFRICA
Algeria Sonatrach Algeries Algeries Refinery RE 60 Bcfd 908 M 2013 Technip
Kenya Kenya Petroleum Refineries Mombasa Mombasa Refinery RE None H
Nigeria NNPC Kogi Kogi Refinery TO 750 bpd 23000 E
ASIA/PACIFIC
Australia Santos\PETRONAS JV Queensland Bowen-Surat Basin LNG EX 1.5 Mtpy 750 P 2011 Total
China CNPC Anning Kunming Oil Refinery Offsites TO 200 Mbpd 3400 U 2012
China CNPC Anning Kunming Oil Refinery Refinery TO 200 Mbpd 3400 U 2012
China Sinopec Zhanjiang Zhanjiang Refinery EX 350 bpd 7600 U 2014
India HMEL Bhatinda Bhatinda Refinery TO 9 m-tpy 189 U 2011
India Indian Oil Corp Ltd Panipat Panipat Hydrocracker EX 1.7 MMtpy 950 U 2010 UOP EIL EIL
India IOCL Paradeep Paradeep LPG Terminal EX 15 m-tpy 900 P 2012
India HPCL Visakhapatnam Visakh Refinery Refinery TO 300 bpd 10000 U 2014
LATIN AMERICA
Bahamas AES Corp Ocean Cay Ocean Cay Desalination 500 Mgpd P 2012
Bahamas AES Corp Ocean Cay Ocean Cay LPG Terminal None P 2012
Bahamas AES Corp Ocean Cay Ocean Cay Storage, LPG 40 Mm3 P 2012
Brazil Petr Brasileiro SA Pernambuco Pernambuco Refinery TO 230 bpd 12000 U 2011
Colombia Ecopetrol Barrancabermeja Barrancabermeja Hydrogen 19 MMscfd 30 C 2010 Technip KTI Axens Sadeven|Ismocol
Colombia Ecopetrol Barrancabermeja Barrancabermeja Sour Water Stripper 500 gpm 21 C 2010 Tipiel|Axens Tipiel|Axens Sadeven|Ismocol
Colombia Ecopetrol Barrancabermeja Barrancabermeja Treater, Tail Gas 110 t/a 17.4 C 2010 Prosernat Axens Sadeven|Ismocol
Colombia Ecopetrol Cartagena Cartagena Refinery EX 165 Mbpd 2600 P 2013 KBR|KBC|EMRE|UOP|Lummus Technology Technip|CB&I CB&I
Jamaica Petrojam Ltd Kingston Kingston Gas Plant TO 7 Mbpd 1300 F 2013 UOP
Mexico Pemex Minatitlan Minatitlan Refinery RE 240 bpd 318 P 2011 Tecnicas Reunidas Constr N. Odebrecht
Venezuela PDVSA Jose Anzoategui Jose Anzoategui Methanol 850 m-tpy 7 C 2010
MIDDLE ST EA
Qatar Shell Royal Dutch Ras Laffan Ras Laffan Gas to Liquid TO 140 bpd 24000 U 2012 Descon Eng JGC|KBR|McDermott
Saudi Arabia Doosan Ras Az Zawr Ras Az Zawr Desalination TO 228 MIGD 1460 U 2014
Saudi Arabia Saudi Aramco Yanbu Yanbu Hydrogen 262 Mcfd 450 U 2014 Air Liquide
Saudi Arabia Al Rajhi Petrochemical Yanbu Yanbu Silicones 6 t/a 1100 P 2014
Turkey Petkim Petrokimya Hldg Yarimca Yarimca Refinery TO 10 m-tpy 10000 U 2015
UAE Takreer Ruwais Ruwais Refinery EX 1271 bbl 623 P 2014 GS E&C GS E&C
UAE Takreer Ruwais Ruwais Refinery (2) EX 127200 bpsd 3110 E 2014 Bayer|Lummus Technology|Axens|Shaw S&W GS E&C GS E&C
UNITED ATES ST
Illinois Citgo Lemont Lemont Hydrotreater, ULSD 42500 bpd 451 C 2010 Mustang Turner
Indiana Clean Coal Refining Corporation Wabash Wabash Coal Liquefaction (CTL)TO 2.5 m-tpy 3000 S 2011
Texas Arabian American Development Beaumont Beaumont Isomerization EX 600 gpm 1 C 2010
Texas Valero Refining Co Sunray Sunray Steam Methane Reformer EX 30 Scfd 14 U 2012 Technip

See http://www.HydrocarbonProcessing.com/bxsymbols for licensor, engineering and construction companies’ abbreviations,

along with the complete update of the HPI Construction Boxscore.

BOXSCORE DATABASE ONLINE

THE GLOBAL SOURCE
FOR TRACKING HPI
CONSTRUCTION ACTIVITY
For more than 50 years, Hydrocarbon Processing
magazine remains the only source that collects and
maintains data specifically for the HPI community,
publishing up-to-the-minute construction projects
from around the globe with our online product,
Boxscore Database. Updated weekly, our database
helps engineers, contractors and marketing personnel
identify active HPI construction projects around the
world to:
• Generate leads
• Market research
• Track trend analysis
• And, decide future budget planning.
Now, we’ve made our best product even better!
Enhancements include:
• Exporting your search results to Excel so you can
compile your research
• Delivering the latest updated projects directly to
your inbox each week
• Designing customized construction reports for your
company using our 50 years of archived projects.
For a Free 2 -Week Trial, contact Lee Nichols at
+1 (713) 525-4626, Lee.Nichols@GulfPub.com,
or visit www.ConstructionBoxscore.com

Select 156 at www.HydrocarbonProcessing.com/RS

HYDROCARBON PROCESSING NOVEMEBR 2010
I 29
Overpressure risk
never goes away.
That’s why we offer total pressure
relief solutions.

Farris Total Pressure

Relief Solutions Include:
Pressure Relief Valves
PSM Engineering /Audit
Service & Software
Asset Management Services
& Repairs

The Total Solution.

It’s simple. Safety relief valves and services to support your facility’s entire
lifecycle. Pressure relief systems designed right from the start. Pressure relief
valves that help your plant operate safely. Global valve maintenance and repair
support. And PSM software and audit services to ensure your plant operates the
way it was designed to during an overpressure event.
Total Pressure Relief Solutions.

We invite you to see all that we can do for you. http://farris.cwfc.com or +1 440-838-7690

Select 88 at www.HydrocarbonProcessing.com/RS

Stretch in technology and gaps in process safety
for the hydrocarbon industry
Dr. M. Sam Mannan, PE, CSP is a Regents
Professor in the Chemical Engineering Depart-
ment at Texas A&M University and director of
the Mary Kay O’Connor Process Safety Center
at the Texas Engineering Experiment Station.
The mission of the Center is to improve safety
in the chemical process industry by conducting
programs and research activities that promote
safety as second nature for all plant personnel
in their day-to-day activities. Before joining
Texas A&M University, Dr. Mannan was vice president at RMT, Inc., a
nationwide engineering services company.
Dr. Mannan is a registered professional engineer in the states of
Texas and Louisiana and is a Certified Safety Professional. His experi-
ence is wide ranging, covering process design of chemical plants and
refineries, computer simulation of engineering problems, mathemati-
cal modeling, process safety, risk assessment, inherently safer design,
critical infrastructure vulnerability assessment, aerosol modeling, and
reactive and energetic materials assessments.
Dr. Mannan co-authored the Guidelines for Safe Process Operations
and Maintenance, published by the Center for Chemical Process Safety,
American Institute of Chemical Engineers (AIChE). He is the editor of
the 3rd edition, Lees’ Loss Prevention in the Process Industries; he has
published 137 peer-reviewed journal publications, two books, seven
book chapters, 142 proceedings papers, 12 major reports, and 143
technical meeting presentations.
Dr. Mannan is the recipient of numerous awards and recognitions
including the American Institute of Chemical Engineers Service to Soci-
ety Award, the Texas A&M University Association of Former Students’
Distinguished Achievement Award for Teaching, the Texas Engineering
Experiment Station Research Fellow, the Texas A&M University Dwight
Look College of Engineering George Armistead, Jr. ’23 Fellow. In 2007,
he was elected Fellow of the AIChE. In December 2008, the Board
of Regents of Texas A&M University System recognized Dr. Mannan’s
contributions in teaching, research and service by naming him Regents
Professor of Chemical Engineering. Dr. Mannan received his BS degree
in chemical engineering from the Engineering University in Dhaka,
Bangladesh. He obtained his MS and PhD in chemical engineering from
the University of Oklahoma.
Oil and gas are contributing enormously to the quality of our
lives in the 21st century, just as they were throughout the 20th
century. With the economies in the various parts of the world
expanding significantly, more energy sources are required by our
society, and in the previous several decades, people have switched
to offshore deepwater hydrocarbon reservoirs. Besides politics,
current exploration and production are limited by the technol-
ogy, for deeper wells, higher pressure reservoirs or crudes that are
difficult to recover because of higher viscosity.
Innovative offshore technology must be developed to carry
out deepwater production and operations. At the same time,
these hazardous operations (i.e., deeper wells and higher-pressure
reservoirs) are creating new and unique hazards. Along with the
current attention on the Transocean Deepwater Horizon Oil Spill,
the development of advanced technologies to ensure the process
safety and operational reliability of offshore facilities is becoming
extremely important.
HPI VIEWPOINT
Hazards offshore. At present, we face the challenge of how to
prevent or control hazards of deepwater exploration and produc-
tion. Process safety is always an essential part of the oil and gas industry
and a core value that requires continual improvement. Regulations
should be dynamic and ready to be modified based on occurring
industrial issues. Accordingly, the government and offshore opera-
tors should develop comprehensive management programs or regu-
lations that assess process safety and environmental hazards.
It is well known that offshore operations have a very special
environment, involving drilling, production and transport, as
well as emergency response to incidents. Offshore employees are
faced with many different factors that increase their exposure to
injury, such as poor weather conditions, high-pressure operations,
chemicals and confined space. Due to the dangerous nature of off-
shore operations, employees typically have very demanding work
schedules. Research and technology developments for operation in
deep waters and high-pressure reservoirs are urgently needed.
The industry must develop theories, analytical techniques
and technologies to improve the current offshore infrastructures
from all sources of failure, including design, operations, manage-
ment, natural disasters and intentional acts such as terrorism.
The research should focus on theories and techniques that apply
to the many process safety issues, such as structural integrity, lay-
ers of protection, offgas handling, drilling, risk assessment and
consequence analysis, human error and safety culture. Test beds
may include processing facilities and complex structures within
the offshore infrastructure, transportation vehicles (e.g., ship,
helicopter) and the marine environment.
The goals of offshore safety research include integration of
the concepts of process safety into the design and operation of
offshore platforms and to use this to improve safety performance
such that the unit/process is not vulnerable to certain failures.
Also, by identifying aspects of the system that are vulnerable (not
resilient), the speed and efficiency of response to failures could be
improved. Some relevant areas of research include:
Structural integrity of risers/pipelines. Risers and pipelines are
subjected to different types of corrosion due to continuous loadings
(fatigues) and their exposure to the marine environment. Research is
needed to study corrosion behavior underneath the coating layer, to
assess the integrity of particular structures and to develop coating/
corrosion assessment criteria for service under extreme conditions.
Layers of protection. Many US offshore rigs are equipped
with blowout preventer (BOP) casing shear rams to seal off an oil
or natural gas well being drilled or worked on. However, the BOP
is currently the only layer of protection within the system and it is
vulnerable to single-point failure. Thus, research efforts are needed
for developing/identifying multiple layers of protection.
Risk assessment and consequence analysis using CFD.
Computational fluid dynamics (CFD) has gained widespread
recognition as a powerful tool for risk assessment and consequence
analysis. The very nature of infrequent and highly diverse disasters
can make prediction of the likelihood and consequences of such
events very difficult. The multitude of systems and structures
HYDROCARBON PROCESSING NOVEMBER 2010
I 31
HPI VIEWPOINT
involved requires a very broad multidisciplinary team to under-
stand, evaluate, compare and plan for disastrous failures. Research
is needed with regard to fluid-structure interaction (vibrations of
risers, motions of floating platforms), flow around vessel hulls in
the presence of current and wind forces, wave loads (slam and
impact), tank sloshing and BOP (impact) facility siting.
Recovery of H2S. Hydrogen sulfide (H2S) is corrosive for
carbon steels used in offshore structures and can lead to nervous
disorders and acid rain. To remove H2S, a process called gas sweet-
ening and amine sweetening has been used for onshore processing
facilities. This process removes H2S from the feed and redirects it
to other processes in the plant, where it is converted into elemen-
tal sulfur or sulfuric acid. However, the present process is limited
by the operating conditions and requires additional units to be
built, which makes it insufficient for the offshore facilities. Given
the compact size of the offshore structures, there is a need to opti-
mize the gas-recovery process for offshore facilities.
Resilient operation of deepwater drilling. During
drilling, all materials drilled out need to be removed, i.e., trans-
ported to the surface, a process that is referred to as hole cleaning.
Often, some material remains in the well. Due to the number of
parameters influencing hole cleaning and the complex mecha-
nisms involved, the phenomenon has not yet been fully under-
stood. Integration of the concept of resilience to the drilling and
hole cleaning processes is needed.
Human error and safety culture. Both human error and
safety culture have been identified as contributing causes in indus-
trial accidents, including offshore facilities. Assessment of the
safety culture in every aspect of work in the organization, and
development of a systematic approach to apply safety culture in
the organization will lead to reduction of human errors.
Spills clean-up. Mechanical containment should be the pri-
mary line of defense against oil spills. However, this method is
not effective in clearing a large spill area. Chemical and biological
methods can be used in conjunction with mechanical means for
containing and cleaning up oil spills. Dispersing agents are the
most useful in helping to keep oil from reaching shorelines and
other sensitive habitats. Biological agents have the potential to
assist recovery in sensitive areas such as shorelines, marshes and
wetlands. In this regard, research for new dispersants, including
nano-surfactant technologies, is needed.
Meeting the challenge. Deeper wells, higher pressure reser-
voirs and crudes with higher viscosity have introduced hazardous
operations in offshore facilities, thus creating new and unique haz-
ards. Efforts should be made to expanding the current focal point
from that of drilling and utilization of BOPs to include a more
encompassing proactive prevention program requirement for all
offshore operations. There are many possible ways in which disas-
ters equal to or greater than Deepwater Horizon could occur in off-
shore operations. Comprehensive programs that include inherent
safety considerations, multiple layers of protection, consideration
of human factors, analysis of worst-case scenarios and emergency
response planning are needed. Finally, major efforts that explore
new technologies in various areas should be a high priority. HP

One level of technical

support: Superior.
Five licenses or 500, when you switch to CHEMCAD 6.3, you’ll
get the same best-in-class support Chemstations offers all of
our users. No wading through several levels of representatives
to get an answer. Every time you contact us, you’ll get the
answers you need, quickly and accurately, from a support
representative who is an expert in chemical engineering
simulation software. And with even more features and
enhancements to advance your engineering projects further,
Version 6.3 is the best reason yet to try CHEMCAD and judge
our technical support for yourself.

Make the switch to CHEMCAD today.

Try it free* at chemstations.com/switch63HP
or call 1.800.CHEMCAD for details.
*Certain restrictions apply.

Engineering advanced
© 2010 Chemstations, Inc. All rights reserved. | CMS-22-1 11/10

Select 157 at www.HydrocarbonProcessing.com/RS

32
I NOVEMBER 2010 HYDROCARBON PROCESSING

Better water management:
A crucial and growing requirement
Glen Messina is global business leader, chemi-
cal and monitoring solutions —water and process
technologies for GE Power and Water. In this role,
he is responsible for setting the strategic direction
of the chemical and monitoring solutions business
as well as overseeing global execution. Mr. Messina
leads the commercial and product management
organizations as well as supporting functions. Pre-
viously, he was chief financial officer for water and process technologies
and was responsible for leading the company’s finance function.
Water touches every industry and every sector, impacting lives,
livelihoods, global industries and local economies. But the world
is facing growing water scarcity challenges, both in developed and
developing countries.
Industry accounts for 20% of all water consumed globally.
Refineries and petrochemical facilities, in particular, tend to be
large water users, both for cooling purposes and in production
operations. The hydrocarbon processing industry (HPI) can play a
key role by adopting comprehensive water-management strategies
and water reuse practices on a widespread basis.
As the global economy slowly recovers, refineries can expect to
see a shift in demand for finished petroleum products in mature
markets and growing demand for fuels by developing regions.
Globally, refining facilities will continually seek ways to increase
production and profit margins, while reducing or eliminating
unscheduled shutdowns and the subsequent untimely cost of
cleaning and replacing equipment. Building a robust water-man-
agement strategy should help refineries to run safer, and more
efficiently, and to be more profitable.
Technology as an enabler. From a technology perspective,
water recovery, treatment, recycling and discharge solutions exist
right now not only to enable conservation, but also to improve
the productivity, efficiency, safety and overall economics of water-
intensive industrial processes. Water technology companies offer
technical services, knowledge and practical experience.
For example, optimizing the performance of open recirculat-
ing cooling systems used in refineries and petrochemical plants
represents an enormous opportunity to both lessen the impact on
freshwater supplies, as well as to reduce costs, increase productiv-
ity and protect assets.
These cooling systems rely on chemically treated water to
reduce water consumption, to prevent deposits from compromis-
ing heat transfer in critical process equipment, to mitigate the
effects of corrosion that threatens capital asset life, and to manage
microbial growth, which, in addition to having a negative impact
on production, can adversely affect human health. But the treat-
ment and control process requires constant monitoring and fine-
tuning, which until now has been cumbersome and inefficient.
Advanced technology is altering that status quo. New technol-
ogy gives users the ability to monitor, measure and adjust multiple
chemicals in the cooling water in an online mode and more pre-
cisely than traditional methods. The result is a better-performing
HPI VIEWPOINT
cooling system with significantly lower operating costs, and at the
same time, it saves hundreds of thousands of dollars in annual fresh
water acquisition costs for a large system.
Another area where advanced water-management technology
plays a key role is to enable the increased use of wastewater in
industrial processes. A case in point is a large refinery in southern
Italy, which had to curtail its use of city water during a drought.
If a second source couldn’t be found, the refinery would have
had to shut down. A nearby municipal wastewater plant could
provide the needed volumes, provided the water could be treated
adequately. Mobile membrane-based reverse osmosis (RO) filtra-
tion technology was the solution.
Once the municipal wastewater was run through the refin-
ery’s own clarifier and filtration system, it was further treated by
multiple RO filtration units until it was clean enough for refinery
processes. Because these units were mobile and could be trucked
to the site, they were able to be installed quickly. As drought con-
ditions and demand continued to grow, the flexible nature of the
system made it easy to add another unit without interrupting the
production of treated water or refinery operations.
Intimate technical understanding needed. A deep
technical knowledge of process chemistry is required to solve
many of the difficult water-management issues faced by HPI
facilities. For example, a Gulf Coast refinery was experiencing
poor dehydration in the processing of heavy Venezuelan crude.
Water carryover and salt content had increased to unacceptable
levels. Although caustic had been injected to control overhead
chlorides, desalter performance was impacting furnace-fuel usage,
unit pressure drop and overhead corrosion control. A compre-
hensive analysis was done, and the ultimate solution was a new
emulsion breaker that was injected into the crude charge.
As time went on, the chemical injection rate was reduced to
about 60% of the historical chemical injection dosages. Total
chlorides were cut by 50% and more consistent corrosion control
in the crude overheads was achieved, even with a reduction in
caustic injection of nearly 30%.
Tomorrow is upon us. Refining crude oil into high-value
fuels and other refined products requires careful coordination and
precision of many complex and expensive systems. Profitability
depends on maintaining the quality and efficiency of these key
process units, controlling operating costs, meeting environmental
requirements and improving efficiencies in every step of the pro-
duction process, while ensuring final product quality.
We have reached the point where the HPI can no longer assume
the availability of adequate, uninterrupted freshwater supplies.
Fluctuating water demand and supply increasingly will become the
norm and the HPI must address growing environmental challenges
and regulations. Partnering with suppliers who understand and
can address concerns with sound water and process solutions will
result in safer, more efficient, more profitable and more environ-
mentally responsive operations. HP
HYDROCARBON PROCESSING NOVEMBER 2010
I 33
Select 55 at www.HydrocarbonProcessing.com/RS
 PLANT SAFETY AND ENVIRONMENT
Water among causes for
storage tank explosion
Reinvestigation uncovers true accident events
M. FERJENCIK and B. JANOVSKY, University of Pardubice, Pardubice, Czech Republic
A
n explosion occurred inside a bitu-
men storage tank. An abrupt inter-
ruption of the purging steam dis-
charge from the tank preceded the accident.
The suspected ignition sources alone would
not be able to cause the observed explo-
sion. An additional impulse must have been
present. Careful analysis of the event led to
the conclusion that the explosion was prob-
ably preceded by water that was introduced
inside the tank. Iron sulfides were present
under the tank roof and combined with
the injection of water (causing the fresh air
suction and atmosphere movement inside
the tank) and this explained the explosion
under actual conditions.
Introduction. Trevor Kletz wrote an
anecdote in which he exaggerated the dan-
gerous properties of water.1 Nevertheless,
he still described water as an extinguish-
ing agent. However, in the accident that
occurred, a situation encountered showed
that water played the opposite role. Even
introducing a relatively small amount of
water into a large bitumen tank seems to
have triggered the accident that resulted in
an explosion and a fire.
This article will present the accident’s
reinvestigation. It is based on a report
that was prepared by a plant investigation
committee. However, the original report
ignored the necessity to explain an abrupt
interruption of purging steam discharge
from the tank which preceded the accident.
The reinvestigation concentrates on this
event and makes it the focal point to shed
light on why the accident occurred.
Installation. The accident occurred more
than five years ago in a plant that processed
heavy-oil hydrocarbons. The event involved
an almost 40-yr-old bitumen storage tank.
The tank was an insulated, carbon-steel,
vertical cylinder with a fixed roof. It was
standing in a concrete emergency sump. The
tank communicated with the surrounding
air via a couple of vents which were placed
at the southern edge of its roof. The total
tank volume was 1,200 m3—height 10 m,
diameter 12.4 m (Fig. 1).
The tank was equipped with openings
and pipe connections. One hatch and seven
pipe connections were in the bottom part of
its wall. There were two connections of the
steam heating system that kept the bitumen
inside the tank hot and liquid. The second
hatch and seven other openings were in
the roof of the tank, and one blind nozzle
was in the upper part of the tank wall. Fig.
2 only shows the roof openings that were
considered to play a role in the accident.
The steam inlet (50-mm diameter) was
located close to two vent nozzles (200-mm
diameter each). A steam pipe with a 25-mm
diameter was inserted into the steam inlet.
Its end piece inside the tank was used for
steam purging. It was less than 1.5-m long;
its mouth was blind and its wall was perfo-
rated. The bitumen circulation inlet pipe
nozzle (125-mm diameter) position was
across the roof at the northwest edge. The
bitumen circulation inlet—a carbon steel
pipe, 100-mm diameter was inserted—was
the newest tank opening. It was welded dur-
ing a scheduled outage of the tank, roughly
six weeks before the accident. The second
most recent opening was two years old.
Chronological order. Operational
records showed that during the last four
days before the accident, the bitumen level
inside the storage tank was relatively low
(maximum 330 cm) and the temperature
was between 150°C–155°C. Steam heat-
ing was operated permanently and steam
purging was inactive. It was in accordance
with operational instructions: steam purg-
SPECIALREPORT
ing was to be used only when the bitumen
temperature inside the tank was higher
than 190°C. More than one day before the
accident, the bitumen level was 65 cm and
at 155°C. These parameters were kept for
more than 24 hr. Conditions inside the
tank started to change again approximately
3 hrs 40 minutes before the accident.
At that time, 120 ton batch of bitumen
was transported into the storage tank. Since
the temperature of the influent bitumen was
higher than 200°C (up to 250°C), steam
purging was commenced. The transport
was finished 40 min before the accident—
the bitumen level was 180 cm and at 205°C.
Steam heating and steam purging were left
in operation; steam pressure was about 4.5
bar and temperature about 140°C. The fill-
ing pipeline was emptied into the storage
tank by pressurized air. All manipulations
connected with the transport were finished
25 min before the accident.
About 25 min later, an explosion
occurred inside the bitumen storage, blow-
ing off the tank roof. After the explosion,
the tank contents started to burn. The roof
fell beside the emergency sump and leaned
against the southwest wall. The explosion
did not damage the tank wall, but a subse-
quent fire destroyed the south-southwestern
part. A layer of coke sediments were prob-
FIG. 1 Bitumen storage tank, front view.
HYDROCARBON PROCESSING NOVEMBER 2010
I 35
SPECIALREPORT
ably burned at relatively high temperatures.
The fire was localized and extinguished in
less than 5 hr. It did not cause material dam-
age outside the bitumen storage tank nor
any fatalities or injuries. Unburned bitumen
had to be placed in barrels and reprocessed.
No anomalies were found in the unburned
liquid during its removal and reprocessing.
Additional observations. The south-
ern part of the tank roof, with vent nozzles,
was within the scope of one camera belong-
ing to the plant security system. Relevant
record analysis provided interesting addi-
tional information about what happened
a few minutes before the accident. Records
confirmed that emptying the filling line was
performed within a 15-min time period,
ending 25 min before the accident—per-
formed in accordance with operational
instructions. Records showed that a per-
manent and stable steam discharge from
tank vents occurred 3 hr 40 min before the
accident and finished abruptly 47 s before
the explosion. During the last 47 s, no
outflows from vents were visible. Also, the
course of the explosion was recorded. The
lifting of the roof was visible, followed by
a rising fireball and flames. Evidently, the
tank roof was blown off by the explosion
of a flammable mixture in the tank atmo-
sphere. The investigations after the accident
showed that the vents were free and that the
steam pipeline to the steam-purging inlet
was open and free, too.
Searching for accident causes. The
fire triangle describes three requirements
that have to be fulfilled for a fire/explo-
sion of a gas mixture: an oxidant, a fuel and
an ignition source.2 Accident causes com-
bine the three requirements. Identifying
the direct accident cause was not possible
without identifying the specific oxidant,
fuel and ignition source that were present
inside the bitumen storage tank.
Steam inlet Bitumen
pipe circulation
Vent inlet pipe
nozzles
S N
FIG. 2 Bitumen storage tank; an aerial
view with selected openings.
36
I NOVEMBER 2010 HYDROCARBON PROCESSING
PLANT SAFETY AND ENVIRONMENT
Oxidant. For at least four days before the
accident, steam purging had been inactive
on the tank. Its gaseous volume commu-
nicated freely with the atmosphere outside
vent openings. Tank space above the liquid
level undoubtedly contained mainly air
at the start of steam purging 3 hr 40 min
before the accident. Sweep steam purging
was used to make the atmosphere of the
tank inert, but it was not able to perfectly
mix the whole gaseous volume when the
liquid level was low.2 Steam is much lighter
than air (and hydrocarbon vapors). The
arrangement of the steam inlet pipe did not
make the steam move into the lower parts
of the tank. Only the upper part of the tank
(about 20% of its total volume, according
to an estimation made by plant personnel)
is believed to have been filled with a steam
blanket. Lower parts of the gaseous volume
probably still contained mainly air during
the explosion.
Large volumes of bitumen in the storage
tank were used from time to time as ter-
minal volume to empty various connected
pipes by air. These emptying operations
brought additional oxidant into the tank.
Fuel. Flammable, gaseous substances had
to be present in the atmosphere inside the
tank in a concentration above the lower
flammability limit for the explosion to
occur. The bitumen itself releases a certain
amount of light hydrocarbons, but mea-
surements indicate that the total content
above the bitumen level is one order below
any conceivable lower flammable limit
(LFL). The bitumen present inside the tank
before the accident was of standard quality;
therefore, the fuel source for the explosion
had to be found elsewhere.
The filling pipeline came into the tank
from a manifold to which pipelines from
a few other storage tanks were also con-
nected. Analyzing operational records
showed that asphalt varnish was pumped
through a pipeline that was connected to
the manifold, more than two days before
the accident. The asphalt varnish repre-
sented a mixture of bitumen (identical
with the stored one) and lacquer diluents.
If a check valve in the pipeline that was
used for pumping asphalt varnish had
not worked properly, a certain amount of
varnish would have entered the manifold.
Insufficiently closing the check valve in its
closing direction is a rather frequent defect
that cannot be excluded.
The liquid amount that would have
entered the manifold in this case might
have easily reached many liters. Undesir-
able liquid containing light hydrocarbons
would then have been transported into the
bitumen storage tank as soon as any of the
pipelines connected to the manifold would
have been emptied into it. Other potential
sources of fuel such as catalytic cracking on
steam heating pipes or steam reforming are
not considered to be probable since tempera-
tures inside the tank were not high enough.
Observed explosion outcomes enabled
estimating the amount of light hydrocar-
bons that had to be present in a flammable
cloud inside the tank. The light hydrocar-
bons originated from lacquer diluents, the
boiling interval was 135°C–220°C, LFL
is 0.8% vol. and upper flammable level
(UFL) is 6.5% vol. To assess the minimum
amount of light hydrocarbons necessary to
have lifted the tank roof off, it was neces-
sary to start with the overpressure, which
could have caused it. Overpressure at 12
kPa is enough pressure to lift a storage
tank roof off.3
The question is, how large does the
flammable cloud have to be if it is capable
of generating 12 kPa of overpressure inside
the tank? From the state equation, it fol-
lows that if the vapor space volume inside
the tank is 989 m 3, then an increase in
the vapor volume should be equal to 117
m3 under normal pressure. This volume
increase is caused by generating hot com-
bustion products. The number of moles
inside the tank should not change during
combustion. Only the temperature differ-
ence between the initial and final states
could cause the volume increase. The sys-
tem’s initial temperature was supposed to
be equal to 478 K. The combustion prod-
ucts’ temperature was estimated to be 1,500
K. This temperature is in accordance that
the flame temperature at the LFL for meth-
ane is 1,498 K and approximately 1,573 K
for other lower paraffinic hydrocarbons.4
Comparing these final and initial states,
an expansion factor equal to 3.14 was
obtained. The equation for the volume of
the explosive mixture capable of producing
the given pressure increase is:
Vexpl + 117 = Vexpl × 3.14.
This results in 54.7 m3 of the explosive
mixture, with a concentration equal to LFL.
The light hydrocarbons may be represented
by C9 fraction with a mean molecular
weight of 148.4 g/mol. Using the molecu-
lar weight, the evaporated flammable vapor
amount is 1.66 kg. It is certainly the lowest
possible amount, not taking into account
the product cooling and venting through
the two vent nozzles. Higher amounts of
 PLANT SAFETY AND ENVIRONMENT SPECIALREPORT

evaporated flammable vapors (e.g., 5 kg and
10 kg) would lead to higher values of the
calculated overpressures (36.2 kPa and 72.4
kPa, respectively). Since these values are
well above the 12 kPa, neither cooling nor
venting through two vent nozzles would
have prevented the roof from lifting off.
Such an amount of flammable vapors could
have easily originated in the asphalt varnish
that entered the tank via the manifold. The
flammable cloud could have been formed
after the temperature increase during the
inflow of hot bitumen between –3:40 and
–0:40 hr. Hydrocarbon vapors are heavier
than air so the operation of sweep steam
purging would not have removed them
from the tank with a low bitumen level.
Fig. 3 illustrates the situation that is
supposed to have been established inside
the tank after adding hot bitumen.
Possible ignition sources. Some igni-
tion sources may include hot work, static
electricity, hot surfaces, pyrophoric iron sul-
fides, pressure (compression ignition), fric-
tion and mechanical sparks, sudden decom-
pression and catalysts.5 Some alternatives
may be excluded immediately.
There was no hot work carried out at
the tank weeks before the accident. A com-
pression or decompression sharp enough to
ignite the flammable vapors is not conceiv-
able under conditions inside the tank. No
moving mechanical parts that would be
able to cause friction or sparks were present
inside the tank.
Movement of nonconductive liquid into
the tank finished at least 25 min before the
accident; hence collection and discharge of
static electricity are not considered to be
probable to ignite the explosion. Hot sur-
faces, in the usual meaning of this term,
were not present inside the tank. However,
a layer of coke sediments were found on the
south–southwestern wall, burning intensely
after the explosion. Suspicion arose—the
coke sediments had been smoldering even
before the accident and they ignited the
flammable cloud. The possible presence of
catalysts (e.g., coke particles with large active
surfaces) in liquid bitumen was considered,
too. The presence of catalytic surfaces could
cause a decrease in the auto-ignition tem-
perature of flammable vapors and lead to
ignition after an induction period.2 Without
a catalyst, auto-ignition of lacquer diluent
vapors is not possible under 240°C.
Pyrophoric iron sulfides form when iron
is exposed to hydrogen sulfide (H2S), or any
other compound that contains sulfur, in an
oxygen-deficient atmosphere. Pyrophoric
iron sulfide may form in heated bitumen
storage tanks as the result of a reaction
between H2S given off from the bitumen
surface and iron in the form of rust on the
tank roof.6 H2S was present inside the tank.
Hence, the area of new bitumen circulation
inlet welds seemed to fulfill all conditions
for pyrophoric iron sulfide formation.
Examining the facts and formu-
lating hypotheses. Three possible
ignition sources were identified: smol-
dering coke, auto-ignition catalysts and
pyrophoric iron sulfides. Examination of
these three hypotheses with known facts is
necessary, and Table 1 represents the fact/
hypothesis matrix.7

TABLE 1. Fact/hypothesis matrix. Legend: (+) compatible with hypothesis; (×) not likely
Temperatures inside tank No anomalies (coke particles, Abrupt interruption of Tank roof fell beside
between 140°C and 205°C, hot spots) were found in steam discharge 47 sec tank in south–
Fact or condition/hypothesis steam blanket under roof unburned bitumen before the explosion southwest direction
Light hydrocarbon vapors + + × ×
ignited by smoldering coke
on south–southwest wall
Light hydrocarbon vapors + × × +
auto-ignited after
induction period
Light hydrocarbon vapors
ignited by pyrophoric iron
sulfides from new weld × + × +

Steam Steam
outlet inlet Air Injection
intake of water
Bitumen circulation inlet pipe

Steam blanket, 140°C

Position of
new weld
Shrinking
of steam
Air with hydrocarbon vapors and air
Fresh air
movement

Hydrocarbon vapors
raised by water
evaporation
Bitumen, 205°C

FIG. 3 Situation inside the tank, 30 min before the accident. FIG. 4 Situation inside the tank during water injection.

HYDROCARBON PROCESSING NOVEMBER 2010

I 37
SPECIALREPORT
TABLE 2. Modified fact/hypothesis matrix. Legend: (+) compatible with hypothesis; (×) not likely
Temperatures inside tank
between 140°C–205°C,
Fact or condition/hypothesis steam blanket under roof
Light hydrocarbon vapors ignited
by smoldering coke on south–
southwest wall after the
introduction of a small amount
of water
Auto-ignition of light hydrocarbon
vapors occurred after the
introduction of a small amount
of water
Light hydrocarbon vapors ignited
by pyrophoric iron sulfides from
new weld after introducing small
amounts of water
Welding of the Return of
bitumen bitumen tank
circulation back into
pipe inlet— operation
approximately Approximately
6 weeks before 4 weeks before
Emptying of
Introduction of
pipe line, 10 m3
waste into
of air introduced
steam inlet pipe
inside the tank
between -0:40
at -00:00:47 sec
and -0.25 hr
FIG. 5 Development of the accident.
Coke smoldering can develop slowly
and gradually, so there would have been no
reason for an abrupt interruption of steam
discharge. If the explosion was ignited by
smoldering coke at the south–southwestern
wall, then the final position of the blown
roof would have been expected on the
opposite side of the tank. Catalyzed auto-
ignition would require the presence of a
catalytic surface on the bitumen. However,
no corresponding anomalies were indicated
in the liquid. Auto-ignition develops slowly
and gradually, thus giving no explanation
for the abrupt interruption of steam dis-
charge. Pyrophoric iron sulfides spontane-
ously ignite after they dry out and come in
contact with air, but there was no indica-
tion that the steam blanket was replaced by
air at new welds. Again, the abrupt inter-
ruption of steam is not compatible with
the hypothesis.
Water injection. None of the consid-
ered ignition sources were able to explain
satisfactorily the abrupt interruption of
steam discharge 47 sec before the explo-
sion. Evidently, the interruption did not
result in any of the conceivable ignition
38
I NOVEMBER 2010 HYDROCARBON PROCESSING
PLANT SAFETY AND ENVIRONMENT
No anomalies (coke particles, Abrupt interruption of steam Roof of the tank fell
hot spots) were found in discharge 47 s before beside tank in south to
unburned bitumen the explosion south-west direction
+ + + ×
+ × + +
+ + + +
Penetration of piece at 4.5 bar. The injection would have
Bitumen level 120 t of bitumen
light hydro- abruptly cooled steam and gases inside the
65 cm, Start of steam added, level
carbons into
temperature purging raised to 180
bitumen
155°C— cm, temp. to
space, causing the gases to shrink. This
inside tank
Approximately
more than At -03:40 hr 205°C between results in under pressure inside the space,
1 day before -3:40 and -0:40 hr
2 days before reversing the flow through the vent nozzles.
The situation is illustrated in Fig. 4.
In Fig. 4, superheated water leaves
Interruption of Penetration of
Spontaneous the holes in the end piece and some of it
ignition of Explosion of air
steam discharge fresh air to evaporates immediately (approximately
pyrophoric and hydrocarbon
pyrophoric iron
iron sulfides, vapors mixture 7.9%). Flash evaporation of water creates
between sulfides at
release of
-00:00:47 and new weld an expanding zone around the end piece
sparks at 00:00:00 sec
00:00:00 sec < 00:00:00
< 00:00:00 sec
inside. The temperature decreases to the
boiling point of water (100°C). The volume
of this zone is relatively small since one liter
of water creates 0.14 m3 of flash evaporated
processes. It resulted from an additional steam. Tiny droplets of boiling water fly
cause and probably contributed to the away from the expanding zone into much
ignition process. Closing or plugging the warmer steam and/or air around and below.
steam pipeline or the outage of the steam The droplets are heated and evaporated. The
supply system would have caused a slow tank atmosphere cools down and shrinks.
decrease of steam discharge, not an abrupt Globally, 1.7 m3 of steam will emerge from
interruption. An event must have occurred 1 L of water. Simple calculations show that,
that caused an immediate pressure decrease in evaporation and balancing temperatures
inside the gaseous volume of the tank. Such in 100 m3 of steam, the steam is cooled by
an event could have been the steam pipeline 18.2°C and shrinks to 4.4 m3. Analogously,
plugging with water. 100 m3 of air would be cooled by 24.8°C
Steam lines need to be equipped by and shrunk to 6.2 m3.
steam traps. A steam trap is a device used to This process leads to a movement inside
discharge condensate and non-condensable the tank atmosphere towards the expanding
gases while not permitting live steam escap- zone around the end piece of the steam-purg-
ing. If the steam trap is not present or if it ing pipe. Possible evaporating droplets on the
fails, then a water plug may form inside the tank walls make this movement even stron-
pipeline. There are indications that a steam ger. If some droplets fell on bitumen liquid
trap was not present in the lower part of the level, they would evaporate and raise hydro-
steam line to the steam purging inlet, so a carbon vapors above the liquid surface.
water plug formation seemed possible. The mixture of water aerosol and cold
If one liter of water had penetrated into steam is relatively heavy and tends to sink
the steam purging line it would have cre- into the air and hydrocarbon mixture.
ated a water plug about 2 m long. The plug Expansion in the steam-aerosol area and
would have been transported into the end turbulences caused by water evaporation
piece of the steam purging pipe. The water make the steam-aerosol area less permeable
would have been injected into the tank’s for downward flowing gas, especially in the
vapor space through tiny holes in the end vicinity of the end piece.
WE ARE 30,000 INNOVATORS DELIVERING SUSTAINABLE ENVIRONMENTAL SOLUTIONS
Jo i n us or h ir e u s. We will s h o w y o u h o w o ur 3 0 , 0 0 0 N o r th A m e r i c an e m ploye e s
are making the world a better place through sustainable solutions. For example, in partnership
with a large utility company, VEOLIA ENVIRONMENTAL SERVICES operates one of the largest
landfill gas-to-high-BTU energy projects in the United States located in Kersey, Pennsylvania.
By piping landfill gas from our Greentree Landfill to an innovative processing plant, VEOLIA
ENVIRONMENTAL SERVICES is providing enough energy to heat approximately 75,000
homes and is reducing greenhouse gas emissions equivalent to over 116,000 tons of
carbon dioxide annually. Learn more at www.veolianor thamerica.com
WATER SERVICES AND TECHNOLOGIES • WASTE SERVICES • ENERGY AND FACILITY MANAGEMENT • PASSENGER TRANSPORTATION

CREATIVE SOLUTIONS FOR OUR ENVIRONMENT

 PLANT SAFETY AND ENVIRONMENT
BORSIG
The estimations confirm that after water
injection, the overall balance of water evap-
oration (with a positive influence on pres-
sure inside the tank) and steam/air shrink-
age (with a substantially higher negative
influence on pressure inside the tank) will
result in air intake through vent nozzles.
Cold and relatively heavy fresh air tends
to sink through the steam pad into the air
and to dilute the air and hydrocarbon mix-
ture. However, there was a zone consist-
ing of steam and aerosol that was denser
than steam and obstructive due to turbu-
lences. This, combined with the sinking
and shrinking movements inside the zone,
would have caused part of the inflowing air
to be sucked in a perpendicular direction
between the zone and the tank roof and
to move along the roof into the peripheral
parts of the tank (Fig. 4). Sinking fresh air
and the steam pad at the southern wall may
have also caused air movement at the oppo-
site side upward against the wall.
Possible ignition sources after
water injection. Three new hypotheses
on possible ignition sources represent the
interactions of sources that were selected
with water injection. The results of the
Leading examination are summarized in Table 2.
If the source were smoldering coke on the
Technology for south–southwest wall after the water injec-
Innovative tion, then the abrupt interruption of the
steam discharge is understandable. How-
Solutions ever, there is no new explanation of why
the tank roof fell in the opposite direction
than would be expected. Similarly, for cata-
Pressure Vessels and lytic auto-ignition the presence of a cata-
Heat Exchangers lyst remains unexplained. The situation has
changed only for pyrophoric iron sulfides.
Reciprocating and
The preceding section explains that fresh air
Centrifugal Compressors could penetrate to a new weld after water
for Process Gases
injection. Hence, conditions for sponta-
Membrane Technology neous ignition of pyrophoric iron sulfides
e.g. Emission Control Systems, would have been fulfilled. There is evidence
Vapour Recovery Units that such an ignition may produce sparks.6
The immediate contact between the sulfide
Industrial Boilers and
Power Plant Technology and the flammable mixture is not necessary.
Consequently, the pyrophoric iron sulfides
Industrial and Power Plant at the new weld might have acted as an igni-
Services
tion source after introducing small amounts
of water into the steam inlet pipe.
Conclusions. The analysis helped com-
BORSIG GROUP plete the time line of events leading to
Egellsstrasse 21 the accident (Fig. 5). Dashed borderlines
D-13507 Berlin/Germany denote the events that cannot be proved.
Phone: +49 (30) 4301-01
Fax: +49 (30) 4301-2236
Probably, the combination of four direct
E-mail: info@borsig.de causes led to the accident: the inefficient
www.borsig.de sweep steam purging of the bitumen storage
tank; the penetration of light hydrocarbons
Select 158 at www.HydrocarbonProcessing.com/RS
into the bitumen inside the tank; the cre-
ation of pyrophoric iron sulfides at the new
weld; and the introduction of water into
the steam inlet pipe. The accident would
never have occurred if a small amount of
water had not been introduced into the
steam purging inlet pipe. It is highly prob-
able that similar situations had occurred in
the tank’s 40-yr history, but the necessary
causes never coincided. Fig. 5 illustrates
a probable multiple-root cause analysis,
which may lead to some lessons learned
from the accident while recommending
proper corrective measures. HP
LITERATURE CITED
1 Kletz, T. A., New Fire-fighting Agent Meets
Opposition, By accident ... a life preventing them in
industry, PFV Publications, London, 2000
2 Crowl, D. A., “Understanding Explosions,”
American Institute of Chemical Engineers, New
York, 2003.
3 Kletz, T. A., “Myths of the Chemical Industry,”
The Institution of Chemical Engineers, Rugby,
1985.
4 Lees, F. P., Loss Prevention in the Process
Industries, Second Edition, Butterworth-
Heinemann, Oxford, 1996.
5 IRP18 Committee, University of Calgary, Canada,
http://www.firesandexplosions.ca/hazards/
ignition sources.php, September 24, 2007.
6 Davie, F. M., T. W. Nolan and S. Hoban, “Study
of Iron Sulfide as a Possible Ignition Source in
the Storage of Heated Bitumen,” Journal of Loss
7
Prevention in the Process Industry, Vol. 6, Issue 3,
pp. 139–143, September 1993.
Center for Chemical Process Safety, “Guidelines for
Investigating Chemical Process Incidents,”
American Institute of Chemical Engineers, New
York, 1992.
Milos Ferjencik studied nuclear
engineering at Prague Technical Uni-
versity and graduated in 1981. He
worked at Nuclear Research Institute,
and in various technical and research
positions. In 1991, Mr. Ferjencik concentrated on chemi-
cal risk analysis. In 1995 he started his own consultancy
profession. Mr. Ferjencik was an independent risk and
reliability consultant and an external teacher at University
of Pardubice. Most recently, Mr. Ferjencik has worked as
a full-time assistant professor of safety engineering at
the University of Pardubice.
Bretislav Janovsky received BS
and PhD degrees from the University
of Pardubice focusing on theory and
technology of explosives. He worked
in various technical and research posi-
tions at the University of Pardubice and a privately held
company, TLP Prague. Dr. Janovsky concentrated on
consequence analysis under risk analysis in the process
industries. He was a full-time assistant and associated
professor of safety engineering at the University of
Pardubice and started his own professional consultant
business in 1995. Today, Dr. Janovsky is the research and
development director at OZM Research Bliznovice and
teaches part-time at the University of Pardubice.
 PLANT SAFETY AND ENVIRONMENT
Consider new analysis for flares
Applying dynamic models in designing safety systems
can reduce capital costs
Z. URBAN, M. MATZOPOULOS and J. MARRIOTT, Process Systems Enterprise Ltd.,
London,UK, and B. MARSHALL, Softbits Consultants Ltd., Medstead, UK
A
pplication of dynamic modeling
for relief system design can sub-
stantially lower capital expenditure
(CAPEX) while simultaneously improving
plant safety. This article considers using
dynamic analysis to two areas: vessel depres-
surization (or “blowdown”) and flare net-
work design. New modeling methods can
accurately quantify relief loads and metal
temperatures to enable informed safety and
CAPEX decision support.
CASE HISTORY 1:
VESSEL DEPRESSURIZATION
Detailed dynamic analysis of the rapid
depressurization (blowdown) of high-pres-
sure vessels is a key element of the safety
analysis of oil and gas facilities and other
high-pressure installations.
Event description. Depressurization of
a vessel usually results in cold gas venting
into the flare system. The cold gas can sig-
nificantly lower the temperatures within the
process equipment metal walls and pipe-
work, as well as the relief system pipework
immediately downstream of the blowdown
valves (BDVs). Low temperatures can lead
to embrittlement of the equipment and
pipework metal walls, and the difference
in temperature between adjacent metal sec-
tions can result in high thermal stresses.
This condition has implications for the
integrity of process vessels, pipework and
sections of the relief system, as well as for
CAPEX. Accurate analysis of likely relief
scenarios is essential to determine:
• Relief loads entering the flare net-
work. For new designs, accurate infor-
mation is needed to achieve an optimal
design that minimizes the piping diam-
eters required to meet Mach number and
back-pressure constraints. Minimizing the
piping sizes also provides benefits in terms
of reduced support infrastructure, which
can be particularly important in the case
of offshore platforms where additional
weight is heavily penalized. For revamps
or expansions to an existing process plant,
accurate data can help determine whether
the current flare system can handle the new
loads acceptably. In either design scenario,
CAPEX savings can be considerable.
• Temperature throughout the pro-
cess and pipework metal walls to identify
areas of potential embrittlement, and where
(and when) unacceptable thermal stresses
are likely to arise. Such information can be
used to mitigate potential problems either
by controlling the relief rates or by rerout-
ing the relief flows.
• Temperature of the relieving “gas”
streams (which may actually contain evap-
orating entrained liquids). This provides
essential information for choosing the
appropriate material of construction for the
critical sections of pipework immediately
downstream of the BDV.
The effects of low temperature can usu-
ally be addressed by using suitable materi-
als of construction. Unfortunately, in some
cases, such materials can be expensive, and
it is highly desirable to minimize their use
without compromising safety consider-
ations. This requires accurate quantifica-
tion of flowrates and temperatures of the
relieving stream, as well as the minimum
temperatures reached in the metal walls.
Complex phenomena. Depressuriza-
tion of a vessel involves a complex set of cou-
pled physical phenomena that must be char-
acterized accurately to understand behavior
and provide suitable design values.
Current depressurization modeling is
often performed with off-the-shelf process
flowsheeting simulators that use an equilib-
rium thermodynamic approach. The latter
provides some indication of the flow and
temperature, but by no means adequately
SPECIALREPORT
describe the complex thermodynamic and
kinetic phenomena occurring as a result of
rapid decreases in pressure. The fact that
multiple phases can form within the vessel,
and that these may not be in equilibrium
with each other or with the vessel wall, can
have a significant effect on both the relief
flows and metal temperatures of the vessel
and relief system pipework.
The sudden decrease in pressure in a
gas-filled vessel results in a rapid change in
the thermodynamic state of the gas within
the vessel. This can result in nucleation
of liquids within the gas bulk to form a
“droplet phase” as shown in Fig. 1. Some
of the nucleated liquid leaves as entrained
droplets in the high-velocity gas exit stream
(Fig. 2). Downstream of the vessel, and as
the pressure further reduces, this exiting
entrained liquid evaporates into the bulk
gas stream, lowering the temperature of
the cold exiting stream even further. This
creates a risk of brittle fracture of the flare
system pipework.
Some proportion of the liquid remain-
ing in the vessel drops to the vessel floor.
Gas to relief system
Nucleating liquid
droplets
FIG. 1 Formation of droplet phase in
blowdown event.
HYDROCARBON PROCESSING NOVEMBER 2010
I 41
SPECIALREPORT PLANT SAFETY AND ENVIRONMENT

Initially, this evaporates instantly due to the The pool boils vigorously, cooling and
warm temperature of the metal it encoun- reducing in size, and, in turn, reducing the
ters. The effect is similar to a drop of water temperature of the metal beneath it. This
falling on a hotplate. However, once the event can lead to significant temperature
metal has cooled sufficiently (typically differences between the metal immediately
after a few tens of seconds), liquid begins below the pool and its surroundings—pre-
pooling (Fig. 2) and forming a continuous senting a very real threat of brittle fracture
liquid phase. and rupture of the vessel base.
The effect of the phenomena can be
seen graphically in Fig. 3, which shows the
Entrained liquid results of depressuring a vessel filled with
droplets leaving light-hydrocarbon gas at 120 bar. In the
with gas initial phase of depressurization, the gas
temperature (black line) drops rapidly. The
temperature of the metal wall in contact
with the gas (green line) begins to drop, but
much more slowly because of the resistance
to heat transfer between between wall and
gas and heat conduction within the wall.
After about 80 seconds, a droplet phase
begins to form throughout the gas. Initially,
droplets in contact with the metal heat up
rapidly and vaporize (red spike). When cool
liquid droplets (at a temperature close to
that of the bulk gas) begin to pool on the
Condensed liquid droplets vessel floor, the liquid temperature increases
pooling and evaporating further above the bulk gas temperature as
the liquid is heated by the metal wall and
FIG. 2 Droplets exiting in the gas stream changes in composition. After a while, the
and forming a continuous liquid gas bulk temperature begins to rise because
phase on the vessel bottom.
of heat influx from the metal wall.
Single-
phase Two-phase
(gas) (gas, droplet)
Three-phase
(gas, droplet,
liquid)
Onset of nucleation Formation of liquid
phase and pooling
300
Metal wall in
290 contact with gas
280
Equilibrium
Temperature, K
The items of most concern are the rap-
idly decreasing temperature of the metal
in contact with the liquid pool (blue line)
and the difference between the tempera-
ture of this metal and the adjacent metal
contacting with the gas (green line). The
metal temperature can be seen to drop
to nearly –30°C, approaching the brittle
fracture temperature for carbon steel. The
temperature difference between vessel floor
and sides rapidly increases to over 20° and
is nearly 40° by the end of the blowdown,
which may give rise to unacceptable stress.
Because of the rapid change of condi-
tions, the three phases coexisting in the
vessel (gas, droplet and a pool of liquid,)
and the vessel walls are not in equilibrium
with each other throughout most of the
blowdown event. In comparison, the dot-
ted line shows the equivalent temperature
curve obtained using an equilibrium model
for the same blowdown, which predicts a
much less severe drop in temperature. This
model fails to identify the most significant
safety-related aspect—the cooling effect of
the liquid on the vessel bottom.
Fig. 4 shows the resulting vessel wall
temperatures and associated thermal stresses
for the vessel vertical walls as color tempera-
ture plots. This information would not be
available without rigorous modeling of the
nonequilibrium mass and energy transfer
between phases. This example describes
just one scenario. Other scenarios may
develop depending on the initial inventory
and state of the material in the vessel. For
example, there may be “bubblet” nucle-
ation in super-critical fluid, rather than the
droplet nucleation described here.
CASE HISTORY 2: FLARE NETWORK
Conventional flare header design tech-
niques use peak relief flows in steady-state
simulation to assess system capacities and
determine back-pressures downstream of
blowdown valves (BDVs) and pressure
safety valves (PSVs), Mach number in the
headers, and radiation at the flare tip.
This steady-state assumption is highly

Metal wall in conservative. While conservative approaches

270 contact with liquid
may be desirable in safety system design,
260 they can nevertheless lead to gross overde-
sign throughout the system. Key areas of
250 Liquid over design include:
240 Bulk gas
Oversized flare header. Sizing the
230 header for the sum of the maximum flows
0 100 200 300 400 500 600 700 800 900 1,000 1,100 1,200
Time, sec takes no account of effects such as:
• System packing, where the gas pres-
FIG. 3 Temperature profiles over the duration of the blowdown event. surizes the available volume in the flare
network
42
I NOVEMBER 2010 HYDROCARBON PROCESSING
 MERICHEM COMPANY

Sweet Solutions.™

The next generation

ti off h
hydrocarbon
treating technologies…under one umbrella
Full-service solutions, upstream and downstream, globally

For 65 years, Merichem Company has been a leader in hydrocarbon sweetening and sulfur extraction
processes. We still are. But there’s more. We’re also a worldwide provider of focused technology,
chemical and service solutions.

Visit our new website at www.merichem.com to see how our customer-tailored, multi-technology
approach to H2S removal, mercaptan removal, caustic management and other solutions can help
you, from beginning to end. And get to know us, all over again.

www.merichem.com

Select 78 at www.HydrocarbonProcessing.com/RS
SPECIALREPORT
• Potential for sequencing of flare
events. For example, depressurization ini-
tiated deliberately by an operator may be
complete well before a fire causes PSVs to
lift. Steady-state peak flow analysis, on the
other hand, assumes that all events occur
simultaneously.
Reducing the peak flows used as the
design basis by judicious analysis can signif-
icantly reduce pipe sizes and materials and
fabrication costs, which can be substantial
for large-diameter headers. Reducing the
size also creates knock-on savings related to
the support structure and flare stack size.
Oversized flare stack. The flare stack
sizing depends on radiation emitted by the
flame, which is a function of the volumetric
gas flowrate through the flare tip. Using
unrealistically high flowrates determined
from peak flows results in an over-long
stack, creating weight problems in offshore
facilities or adding stack support costs (or
unnecessary additional header length) in
onshore facilities. Similarly, a lack of accu-
rate temperature information leads to a wide
span between the minimum and maximum
design temperatures used for gas arriving at
the stack, resulting in unrealistic allowances
for thermal expansion and contraction.
Over-use of expensive alloys.
Although flare system pipework may be in
contact with gas at extremely low tempera-
tures, this typically occurs for a relatively
short duration. The use of steady-state
flows does not consider the duration of
such exposures to low temperature, which
may result in very conservative and expen-
sive application of alloys.
It can be argued that a good flare net-
FIG. 4 4a. Wall temperatures at the end of the blowdown;
4b. Wall thermal stresses from the effects of pooling liquid.
44
I NOVEMBER 2010 HYDROCARBON PROCESSING
PLANT SAFETY AND ENVIRONMENT
work design is one that minimizes capital
expenditure while meeting all safety con-
straints. Overdesign should be avoided
wherever possible.
By making simple dynamic analyses
using data that is mostly already available
in some form, it is often possible to refine
network designs to arrive at systems with
a significantly lower capital cost while
demonstrably meeting safety require-
ments. Similarly, it is often possible to find
additional capacity during retrofits, thus
removing the need for additional capital
expenditures.
Typical examples of where dynamic
analysis can bring significant new informa-
tion that has an impact on capital cost are:
Peak flowrates. The actual relief flow
through any PSV is at the maximum only
for a short period. Using steady-state meth-
ods based on peak flows is equivalent to
making the assumption that all relief flows
start at the same time and go on forever. In
reality, it is often possible to take credit for
staged or staggered relief. Shifting depres-
surization of certain units by a few tens of
seconds can make a significant difference
to the peak flows through the system—an
effect that cannot be represented at all by
steady-state simulation.
Packing. Steady-state approaches make
the implicit assumption that the flare sys-
tem has no volume—what goes in comes
out, instantly. For larger systems, the
impact of relief flows is partially “absorbed”
by pressurization of the flowing lines and
the dead volumes in non-flowing parts of
the system. This “packing” effect can reduce
both the calculated peak back-pressures or
Mach numbers and the peak flows seen at
the flare tip, allowing reduction in header
FIG. 5 Example of a flare network showing active sources.
and tailpipe diameters and flare stack
lengths, respectively. Dynamic simulation
allows this important buffering effect to be
taken into account in the design.
Duration. Equally important, dynamic
simulation can be used to determine the
duration of peak flare loads. Engineering
judgment can then be used to assess the
risks of any infringements. For example,
a 5-second violation of back-pressure or
radiation constraints may well be accept-
able, especially given the capital costs of
oversizing the flare system to avoid such a
contingency.
Temperature. Relief system pipework
that is likely to encounter low temperatures
needs to be constructed from expensive
alloys such as Inconel to avoid the possibil-
ity of embrittlement and consequent frac-
ture. The true extent of pipework that truly
needs to be constructed of such materials is
impossible to gauge with steady-state simu-
lation, as low-temperature flows are con-
sidered to continue forever, ensuring that
calculated metal temperatures reach their
minimum. In reality, such flows may only
last for a few minutes; the thermal inertia of
the pipework metal and heat gain from the
environment prevent the pipework from
reaching the gas temperatures during this
time (a similar effect can be seen in the bulk
gas and metal temperature plots in Fig. 3).
It is frequently possible to reduce the usage
of alloy significantly based on the more
accurate information from the dynamic
analysis. One oil company reported saving
$1.5 million on a single vessel this way.
Flare-stack temperatures. Dynamics
can also help provide a true picture of the
temperature of gas arriving at the flare tip.
Proper calculation of the effect of low-tem-
MERICHEM COMPANY

Problem: Sour Gas

®
Sweetest Solution: LO-CAT
H2S Removal/Recovery
With a wide range of system capabilities, Merichem Company
offers both liquid and solid media desulfurization technologies
to sweeten gas and air streams containing virtually all levels of
hydrogen sulfide or mercaptans.

Our extensive experience includes treatment of natural gas,

amine acid gas, sour water stripper gas, fuel gas, hydrogen
recycle gas and Claus tail gas.

LO-CAT’s flexible technologies and worldwide services are

backed by over three decades of reliability. Join 200+ licensees
in 29 countries on 6 continents meeting today’s compliance
requirements by turning nasty H2S into elemental sulfur —
efficiently, effectively and environmentally sound.

Learn how sweet it is, visit www.merichem.com/LO-CAT

Merichem: A global provider of focused hydrocarbon-treating

technology, chemical and service solutions.

Sweet Solutions.™

P: 847.285.3850 | E: mgtsales@merichem.com | www.merichem.com

Select 79 at www.HydrocarbonProcessing.com/RS
SPECIALREPORT PLANT SAFETY AND ENVIRONMENT

0.30 0.5

0.4

Mach number
Mach number

0.20
0.3

0.2
0.10
0.1

0.00 0.0
0 100 200 300 400 500 600 700 800 900 0 100 200 300 400 500 600 700 800 900
Time, sec Time, sec

FIG. 6 Mach numbers of 24-in. header in a flare system. FIG. 7 Mach numbers of 20-in. header in a flare system using a
staggered blowdown.

TABLE 1. Back-pressures for
24-in. header case
Inlet source Back-pressure, bar
Train_1_DP
Train_2_DP
Train_3_DP
Train_4_DP
Train_5_DP
TABLE 2. Back-pressures for
20-in. header case
Inlet source Back-pressure, bar
Train_1_DP
Train_2_DP
Train_3_DP
Train_4_DP
Train_5_DP
perature gas over a short time taking into
account metal-thermal inertia and ambi-
ent heating provides much more accurate
minimum and maximum design tempera-
tures, allowing the designer to make sen-
sible decisions on stack length and support
mechanisms.
Example. Consider the flare system
shown in Fig. 5, where the header sizes are
set primarily by a depressuring scenario
from five units simultaneously, as high-
lighted in the figure. One of the key ques-
tions is the size of the long main-header
pipe leading to the flare stack. Typically,
such a system is designed by working back
from the flare tip, sizing all the lines based
on velocity constraints until reaching the
relief valves, and then confirming that
other constraints, such as back-pressure
constraints at the relief valves and limits on
noise, are not violated.
In this study, a system has been designed
using steady-state techniques using a veloc-
46
I NOVEMBER 2010 HYDROCARBON PROCESSING
ity heuristic requiring a Mach number
between 0.25 and 0.35 in the main lines.
A conventional steady-state approach cal-
1.63
culates the Mach number using the sum of
the peak flowrates. The maximum Mach
1.75
number at the pipe outlet is represented
1.63 by the dotted line in Fig. 6. At 0.29, this
1.79 is well within the 0.25–0.35 range. The
1.68 back-pressures at the five blowdown valve
sources are listed in Table 1; these are well
below the limit.
For illustration, it is useful to do a
“pseudo-dynamic” run, using relief flow
curves but taking no account of the flare
2.23 system volume. This shows a Mach num-
2.32 ber profile over time that has the charac-
2.23 teristic sharp-peaked shape of relief flow
2.34
curves (red line in Fig. 6). As expected,
the peak Mach number from this run is
2.27
the same as for the steady-state case, at the
sum of the individual peak flows. Although
this case adds no new information to the
design, it does provide some indication of
the length of the blowdown event, allow-
ing judgment to be applied in the case of
constraint violation.
If a full dynamic simulation is done,
taking the volume of the system (both for
active and inactive branches) into account,
it can be seen that the effect of flare sys-
tem packing significantly reduces the peak
Mach number observed, to about 0.25.
It is evident from these results that there
may be potential to reduce the diameter of
the 24-in. header, as the Mach number is
nowhere near its limit.
A new series of calculations is performed
with a 20-in. header diameter to see the
effect of reducing the flare system line sizes.
As expected, the Mach number obtained
using steady-state peak flows (0.4) violates
the system design constraints, indicating
that the design is not viable (Fig. 7, dotted
line). The corresponding pseudo-dynamic
case shows (Fig. 7, red line) that the value
is out of range for about three minutes,
which is also unacceptable. However, a full
dynamic simulation taking into account
line packing shows that the peak Mach
number is within the 0.35 limit (Fig. 7,
black line); the back-pressures (Table 2)
remain well within the limits. The added
information provided by the dynamic
simulation thus indicates that the design
is indeed viable.
If further mitigation is required, it is pos-
sible to investigate the dynamic effects of
staggering the depressurization, so that units
depressure in sequence. The green line in
Fig. 7 shows the effect on Mach number of
delaying the blowdown of Unit 2 by several
minutes. Similar approaches can be applied
to retrofit cases, often demonstrating that
it is possible to accommodate additional
sources in an existing flare system that is
ostensibly operating close to its limits.
Conclusion. The dynamic simulation
capabilities of modern software tools
provide a number of options for analyz-
ing both the depressurization event—to
determine accurate relief flows and fluid
thermodynamic conditions—and the flare
header design itself. This enables engineers
to design systems that comply with safety
guidelines based on a much more realistic
representation of behavior than traditional
methods allow, and, at the same time, to
identify opportunities for significant capi-
tal savings.
In the case of depressurization, rigorous
dynamic simulation identifies potentially
dangerous situations. For new flare system
designs, it can lead—among other benefits—
to a reduction in header size, resulting in
significant capital savings. For existing head-
ers, it provides a means to establish whether
there is sufficient capacity to accommodate
new sources, thereby avoiding the need for
a new header and flare. HP
 PLANT SAFETY

BIBLIOGRAPHY
Haque, M. A., S. M. Richardson, and G. Saville,
“Blowdown of Pressure Vessels. I—Computer
Model,” Transactions of the Institute of Chemical MERICHEM COMPANY
Engineers Part B: Process Safety Environmental
Protection, 70(BI), 1, 1992.
Mafgerefteh, H. and S. M. A. Wong, “A numeri-
cal blowdown simulation incorporating cubic equa-
tions of state,” Computer Chemical Engineering, Vol.
23, p. 1309, 1999.
Gruber, D., D.-U. Leipnitz, P. Sethuraman, M. A.
Alos, J. M. Nougues and M. Brodkorb, “Are there
alternatives to an expensive overhaul of a bottle-
necked flare system?” Petroleum Technology Quarterly
Q1, 2010.
Chen, F. F. K., R. A. Jentz and D. G. Williams,
Problem: High jet fuel
“Flare System Design: A Case for Dynamic
Simulation,” Offshore Technology Conference, May processing costs
Opportunity: Lower CAPEX
4–7, 1992, Houston.
Goyal, R. K. and E. G. Al-Ansari, “Emergency
Shutdown devices and relief system sizing and design

and OPEX by avoiding solid

in oil refineries,” Hydrocarbon World, Vol. 4, No. 1,
Touch Briefings, 2009.
Speranza, A. and A. Terenzi, Blowdown of
Hydrocarbons pressure vessel with partial phase separa-
tion, Series of Advances in Mathematics, available
from http://www.i2t3.unifi.it/upload/file/Articoli/
bed media
animp_2004.pdf, 2005.
Szczepanski, R., “Simulation programs for blow- Solution: Announcing
down of pressure vessels,” IChemE SONG Meeting,
1994.
next-generation MERICAT™ J
Zbigniew Urban is a principal consultant at Process
Systems Enterprise Ltd. He has 30 years of experience
in the application of high-fidelity modeling to complex MERICAT J uses FIBER-FILM®
industrial processes, including areas such as reaction
patented technology and
engineering, gas-liquid separation, industrial crystal-
lization and depressurization to fuel -cell component proprietary solution to sweeten
and system development. Mr. Urban is responsible for mercaptans like no other without
development of PSE’s novel modeling technology, and
is an acknowledged leader in these fields. He has an
the need for solid bed media. See
MS degree in chemical engineering from the Warsaw how next-generation technology,
University of Technology. with lower capital and operating
costs and a smaller plant footprint
Mark Matzopoulos is chief operating officer of
Process Systems Enterprise Ltd. He has 30 years of experi- can make more money for you.
ence in the development and application of process sim-
ulation and modeling tools in steady state and dynamic
simulation, operator training, control system checkout
Finding the right treating solution
and plant optimization in the oil and gas, refining and to remove impurities from hydrocarbon streams is challenging
chemical industries. He has a chemical engineering for any refiner. But Merichem’s decades of experience and
degree from the University of Cape Town, South Africa,
and was a member of the winning team for the Royal commitment to innovation means treating gaseous and liquid
Academy of Engineering MacRobert Award 2007. hydrocarbons can be efficient, economical and clean.

James Marriott is head of applications engineering Mercaptans in jet fuel/kerosene are bad.
at Process Systems Enterprise Ltd. He is responsible for
safety-business development and has 10 years of expe- Learn what’s sweet at
rience in industrial application modeling. Mr. Marriott www.merichem.com/MERICAT-J
holds a degree in chemical engineering from University
College London, England, and a PhD in modeling of
separation systems. Merichem: A global provider of focused
technology, chemical and service solutions.
Brian Marshall has over 33 years of oil and gas
experience, including 10 years of process design in
onshore and offshore oil/gas and condensate produc- Sweet Solutions.™
tion facilities, 2 years of commissioning and operation
of major LNG facilities and 25 years in flare-system
design and operation. He is the managing director of
Softbits Consultants, a specialist supplier of process
simulation software and consultancy for flare system
design. Mr. Marshall is a graduate of Nottingham Uni-
versity, a former technical chairman of GPA Europe and P: 713.428.5000 | E: mptsales@merichem.com | www.merichem.com
a Fellow of the IChemE.

Select 159 at www.HydrocarbonProcessing.com/RS

47
Lewis Pumps Casts a Long Shadow
With installations in over 100 countries
worldwide, Lewis vertical molten sulphur
and sulphuric acid pumps are the
first choice of sulphur recovery and
acid plant operators everywhere.

Standard replacement parts are

continuously available for all
Lewis pumps. In case of an
emergency you will have what
you need within 72 hours to any
major airport worldwide. Guaranteed!

Customers in over 100 countries

can’t be wrong.

SOMEWHERE, A LEWIS PUMP IS OPERATING RIGHT NOW...

CHAS. S. LEWIS & CO., INC. • 8625 Grant Road • St. Louis, Missouri 63123 U.S.A.
Telephone: (314) 843-4437 • Telefax: (314) 843-7964 • E-mail: sales@lewispumps.com
Select 67 at www.HydrocarbonProcessing.com/RS
 PLANT SAFETY AND ENVIRONMENT SPECIALREPORT

Designing the correct

pressure-relieving system
Use these relief rate calculations for gas thermal
expansion as a cause for overpressure
S. RAHIMI MOFRAD, Petrofac Engineering and Construction, Sharjah, UAE

T
he first step in designing a pressure-relieving system is to
specify all possible causes of overpressure that is applicable
to different equipment. The American Petroleum Institute
(API) standard 5211 introduces the various emergency cases and
general guidelines for calculating relief rates. While liquid thermal
expansion is widely recognized as causing overpressure, gas ther-
mal expansion is rarely taken as a credible cause of overpressure.
Standard guidelines may be absent or it may be believed that gas
thermal expansion has negligible contribution to overpressure.
As a general practice for liquid thermal expansion, thermal relief
valves (TRVs) are normally provided for equipment operating full
of liquid. These can be blocked-in between inlet and outlet valves,
where sufficient heat may be applied to fluid, increasing the pres-
sure above the equipment design pressure.2 The relief requirement
of a vessel under fire containing only gas (unwetted vessel) was
developed by API 521 and other reference materials.3 This article
will present cases where gas thermal expansion is caused by heat
sources other than external fire—including process hot stream,
solar radiation and ambient temperature variation.
Gas expansion fundamentals. Gas thermal expansion
occurs in all equipment and piping with the following conditions:
• Line or equipment is isolated for operational or emergency
purposes
• A heat source with temperature higher than gas temperature
exists.
For the described system, a relief valve is needed if:
• Pressure increased since the temperature increased, exceed-
ing the system design pressure
• No other overpressure protection device on the system, no
provision to prevent gas tight blockage such as a locked open (LO)
valve, leaking check valve or venting procedure before closing
isolation valves.2
Figs. 1A and 1B show the pressure rise of blocked air, based on
the results of the Soave–Redlich–Kwong (SRK)4 equation of state
(EOS) for different pressures and temperatures. As shown in Fig.
1A, the blocked-in gas pressure rise due to temperature changes
(ΔP/ΔT), is not significant in low to medium pressure applica-
tions. However, for high pressure gases it can reach 2 bar/°C (Fig.
1B). The values given in Figs. 1A and 1B may be used to estimate
the potential pressure rise for other gases, but for gas mixtures
with a wide composition range, use the suitable EOS.
Figs. 1A and 1B illustrate that gas thermal expansion should be
considered as causing overpressure in the following systems:

0.10 2.5

0.09 1 bara 100 bara

5 bara 200 bara
10 bara 300 bara
0.08 20 bara 2.0 500 bara
30 bara
0.07
ΔP/ΔT, bar/1°C

ΔP/ΔT, bar/1°C

0.06 1.5

0.05

0.04 1.0

0.03

0.02 0.5

0.01
0.00 0.00
-50 0 50 100 150 200 250 300 -50 0 50 100 150 200 250 300
Temperature, °C Temperature, °C

FIG. 1A Gas expansion for low to medium pressure. FIG. 1B Gas expansion for medium to high pressure.

HYDROCARBON PROCESSING NOVEMBER 2010

I 49
SPECIALREPORT PLANT SAFETY AND ENVIRONMENT

• Systems with high operating pressure where ΔP/ΔT is PD
PN

significant. This means that even small temperature changes may
cause catastrophic piping or equipment failure containing dense
gas. Example, a 300-bar air cylinder stored at 23°C, the pressure
will increase 1.3 bar per each degree centigrade, as shown in Fig.
1B. Therefore, if the air condition fails and room temperature
increases to 40°C, cylinder pressure can reach 322 bar. Another
example is a high-pressure pipeline with a possible blockage, along
with exposure to sun radiation.
• System with high differential temperature between heat
source and blocked gas. A 20-bar cold fuel gas is superheated in
a gas–gas heat exchanger using hot flue gas at 400°C. If the fuel
gas is blocked, it can reach the hot gas operating temperature.
Although the gas expansion is small at 20 bar, the high differential
temperature can easily increase the blocked gas to pressure beyond
the design pressure.
Relief rate calculation. There is not a common approach
for gas thermal expansion relief studies that is applicable to all
systems. These systems should be reviewed case by case. What is
common among different systems is that pressure can be main-
tained at a safe level if the excess mass is released. Writing a mass
balance equation for blocked gas gives the required relief rate:

dm(t )
W (t )= (1)
dt
Assuming:
• Since the gas composition does not change during relief, gas
molecular weight is constant.
Hence, m(t) = MW n(t)
• Enough heat is supplied to keep the blocked gas at relieving
pressure during relief (independent of time (P = PR ))
• n = PV/RT predicts gas thermodynamic behavior.
Eq. 1 could be rewritten as:
PR V MW d (1 /T )
W (t ) = (2)
R dt
Rearranging Eq. 2 results in Eq. 3, and can be solved if the
temperature variation with time is known.
P V MW dT
W (t ) = R (3)
RT 2 dt
Initial checking. An appropriate safeguard for gas thermal
expansion is equipment design pressure. Before starting the relief
rate calculation, it should be checked at which temperature the
design pressure will be reached using Eq. 4.
= Fig. 1value at PN andTN ,ave (4)
T
TN ,ave
A simplified form of the ideal gas law, T = TN,ave PD /PN, con-
sidering its limitations and applications can also be used.
If the calculated temperature, T, is higher than the heat source
temperature, no pressure safety valve is needed. In other words,
the pressure of blocked gas does not reach relieving pressure even
if the temperature increases from the initial temperature to the
heat source temperature. If not, there is a potential for the blocked
gas pressure to exceed the design pressure. The calculated tempera-
ture is the temperature where a relief valve opens for the first time
(the initial relieving temperature).
Relief rate calculation procedure. The following section
introduces a very simple procedure to define the blocked-gas tem-
perature as a function of time and solves Eq. 3 for a gas-gas heat
exchanger using a numerical solution. This method may result in
10%–20% overdesign on flow (required area). It’s also acceptable
compared to the overdesign associated with the relief valve selec-
tion procedure—selecting a pressure safety valve (PSV) among
standard orifices usually much larger than the required area.
Step 1. Assume constant temperature intervals, ΔT, between
the initial relieving temperature and the heat source temperature
and calculate T(t+Δt):
T (t +
t ) =T (t ) +
T (5)
Step 2. Calculate the average heat transfer rate between heat
source and trapped gas by using Eq. 6:
Q(t ) =U (t ) A (TS
T (t ))
(6)
Step 3. Calculate the heat required to increase the tempera-
ture of trapped gas from T(t) to T(t+Δt) by using Eq. 7, which
ignores part of the heat that is consumed by the gas container
(heat-exchanger metal).
q(t ) = n (t ) MWC (t ) T (t + t ) T (t ) (7)
P
Step 4. Calculate the time required to increase the temperature
of trapped gas from T(t) to T(t+Δt) using Eq. 8.
q (t )

t = (8)
Q (t )
Step 5. Calculate the relief rate using Eq. 9.
PR V MW (T (t + t ) T (t ))
W (t ) = (9)
RT (t )2 t

Step 6. Calculate the new number of trapped gas moles using

TABLE 1. Gas condition and composition
Eq. 10 for the next calculation stage. Go back to Step 1.
Parameter Value
n (t + t ) = n (t )
W (t ) t (10)
Methane (mole %) 0.84
Ethane (mole %) 0.08
Case study. Table 1 shows natural gas with conditions and
compositions that are superheated from 100°C–150°C in a gas-
Propane (mole %) 0.04
gas heat exchanger using 400°C high-pressure steam. The heat-
i-butane (mole %) 0.02 exchanger tube inside diameter, total heat transfer area and total
n-butane (mole %) 0.02 tube volume are 1 in., 100 m2 and 1.5 m3 respectively.
MW 19.97 According to Fig. 1A, the pressure rise of trapped gas at an
PN (barg) 10 average temperature of 125°C and 10 barg is 0.0289 bar/°C.
PD (barg) 12 Therefore, the minimum heat source temperature to increase the
blocked-gas pressure from operating to design pressure is:
50
I NOVEMBER 2010 HYDROCARBON PROCESSING
SPECIALREPORT PLANT SAFETY AND ENVIRONMENT

TABLE 2. Relief rate calculation

T (°C) Cp* (kJ/kgK) ␮* (cP) ␳* (kg/m3) k* (J/s m K) Gr (—) Pr (—) Nu (—) U (J/sm2K) Q (J/s) m (kg) q (J) ⌬t (s) W (kg/hr)
235.7 2.84 1.75E-02 6.74 6.42E-02 7.32E+06 0.775 11.7 30.0 493205 10.10 470582 0.95 1241.1
252.1 2.90 1.79E-02 6.53 6.70E-02 5.71E+06 0.776 10.8 29.1 429724 9.78 464857 1.08 1027.3
268.5 2.96 1.83E-02 6.32 6.98E-02 4.42E+06 0.777 10.0 28.0 368505 9.47 462289 1.25 838.1
285.0 3.02 1.87E-02 6.13 7.27E-02 3.38E+06 0.778 9.2 26.9 309314 9.17 454386 1.47 669.9
301.4 3.08 1.91E-02 5.95 7.56E-02 2.54E+06 0.779 8.5 25.7 253048 8.90 449576 1.78 522.8
317.8 3.14 1.95E-02 5.78 7.86E-02 1.87E+06 0.780 7.7 24.3 199593 8.64 447697 2.24 393.8
334.3 3.20 1.99E-02 5.62 8.15E-02 1.32E+06 0.781 7.0 22.7 149015 8.40 440545 2.96 281.0
350.7 3.26 2.03E-02 5.47 8.45E-02 8.82E+05 0.782 6.1 20.8 102460 8.17 436291 4.26 185.0
367.1 3.32 2.06E-02 5.33 8.75E-02 5.25E+05 0.782 5.2 18.4 60452 7.95 434803 7.19 104.6
383.6 3.37 2.10E-02 5.19 9.06E-02 2.34E+05 0.783 4.1 14.9 24373 7.74 428127 17.6 40.5
400.0 3.43 2.14E-02 5.07 9.36E-02 0.00E+00 0.783 0.0 0.0 0 7.54 — ∞ 0.0
*Input from simulation software

Heat transfer background PD PN

= 0.0289 bar / °C
T 125°C
Ignoring the effect of heat radiation, overall heat transfer
coefficient for the tube side of the heat exchanger is calculated Then:
from the following relation: 12 10
T = +125 =194.2°C
0.0289
1 1 1 1
= + + (11)
U hi kw  Aw   Ao  Since the temperature of HP steam is higher than 194.2°C,
ho  

x  Ai   Ai  it will over-pressurize the cold side and a relief valve is required
for gas thermal expansion if the system is blocked-in. Simulation
software was used to obtain the physical properties of the gas at
From the heat transfer point of view, the main difference each temperature interval. Considering relieving pressure of 13.2
between the operating and blocked-in conditions is that the barg (corresponding to a relieving temperature of 235.7°C), the
heat transfer mechanism inside the tube changes from force relief rate was calculated (Table 2).
convection to free convection. The following assumption is In this example, the following points were observed:
applicable to blocked-in conditions: • As the blocked-gas temperature increases, the differential
temperature decreases. This results in reducing the overall heat
kw transfer coefficient and, subsequently, the relief rate.
hi  h (12) • The maximum relief load takes place at the first interval

x o
when the heat transfer rate is high.
• The time between two subsequent PSV openings (Δt) is
Substituting Eq. 12 in Eq. 11 gives U≈hi. initially so short that it can be assumed as continuous relieving.
There are many correlations for calculating the free convec- The PSV will chatter as trapped gas temperature approaches heat
tion heat transfer coefficient, hi, inside enclosed space. The fol- source temperature.
lowing relation can be used for estimating this parameter when • Unlike liquid thermal expansion where a ¾ in. x 1 in. ther-
6 x 106 < Gr Pr < 108.5 mal relief valve will normally be sufficient, preliminary PSV sizing
showed that a 1½ in. x 2 in. PSV with an F designation is needed
Nu = 0.104 Gr0.305 Pr0.389 for this case.
• When the gas temperature reaches 194.2°C, the relief valve
where: opens for the first time. If the heat transfer rate is high, the relief
valve will remain open, otherwise, it will close until the pressure
buildup is sufficient to reopen the relief valve again.
g (TS T ) D 3 μC p The typical trends of blocked-gas parameters for this case are
Gr = Pr =
2 k illustrated in Fig. 2 (Case 1). In case of low heat flux, the relief
valve opens and closes repeatedly to release the excess pressure, as
shown in Fig. 2 (Case 2).
hi D μ Gas thermal expansion can cause system over-pressurization
Nu = =
k in particular conditions that were discussed. Ignoring this case
may result in a system mechanical failure. A ¾ in. x 1 in ther-
mal relief valve recommended by API-521 may not be suffi-
52
I NOVEMBER 2010 HYDROCARBON PROCESSING
AD: www.graficadueprint.com
© 2010 Costacurta S.p.A.-VICO

SINCE 1921...
AND WE
STILL LOVE IT

For more than eighty years, we at Costacurta have

been constantly and resolutely committed to the
development and manufacture of special steel wire and
plate components used in many different industrial
processes. Every day at Costacurta, we work to
improve the quality of our products and services and the
safety of all our collaborators, paying ever-greater
attention to the protection of the environment.
Within the wide range of Costacurta products you will
also find some, described below, that are used specifically
in the oil, petrochemical and chemical industries:
- RADIAL FLOW AND DOWN FLOW REACTOR INTERNALS;
- GAS-LIQUID AND LIQUID-LIQUID SEPARATORS;
- ARMOURING OF REFRACTORY, ANTI-ABRASIVE AND
ANTI-CORROSIVE LININGS.
For more information visit our website or contact the
division 'C' components for the oil, petrochemical
and chemical industries at tcsc@costacurta.it.
Gas-liquid and liquid-liquid separators

Costacurta S.p.A.-VICO Management systems

via Grazioli, 30
20161 Milano, Italy
tel. +39 02.66.20.20.66
fax: +39 02.66.20.20.99
www.costacurta.it
certified by LRQA:
ISO 9001:2008
ISO 14001:2004
OHSAS 18001:2007
Select 57 at www.HydrocarbonProcessing.com/RS
SPECIALREPORT PLANT SAFETY AND ENVIRONMENT

cient for this case. Therefore, the relief rate has to be calculated NOMENCLATURE
according to the system dimension and heat transfer rate from A Heat exchanger surface area, m2
Cp Gas heat capacity, kJ/kg K
the heat source to the blocked-in gas while an adequate PSV D Tube diameter, m
size needs to be utilized. In absence of any standard addressing g Accleration of gravity, 9.81 m/s2
this case, different systems should be reviewed on a case-by- Gr Grashof number, dimensionless
case basis. HP h Heat transfer coefficient, J/s m2 K
k Thermal conductivity, J/s m K
LITERATURE CITED m Mass of trapped gas, kg
1 “Pressure-relieving and depressuring systems,” API RP 521, 5th edition, MW Gas molecular weight, kg/kgmol
January 2007. n Mole of trapped gas, kgmol
2 Norouzi, S. and S. Rahimi Mofrad, “What you should know about liquid Nu Nusselt number, dimensionless
thermal expansion,” Hydrocarbon Processing, November 2008. P Pressure, bara
3 Rahimi Mofrad, S. and S. Norouzi, “Designing for pressure release due to a Pr Prandtl number, dimensionless
fire—Part 1,” Hydrocarbon Processing, November 2007. Q Total heat transfer rate, J/sec
4 Smith, J. M. and H. C. Van Ness, “Introduction to Chemical Engineering q Heat content of trapped gas, J
Thermodynamics,” Fourth Edition, McGraw Hill. R Gas constant, 8,314 bara m3/kgmol K
5 Cheremisinoff, N. P., Heat Transfer Handbook, Gulf Publishing Company,
Ra Rayleigh number, dimensionless
Houston, Texas, 2003.
TS Heat source temperature, K
T Blocked-gas temperature, K
t Time, sec
U Overall heat transfer coefficient, J/s m2 K
Case 1: High heat flux V Trapped gas volume, m3
Case 2: Low heat flux W Relief rate, kg/hr
T=Ts Z Gas compressibility factor
␤ Cubical expansion coefficient, 1/K
Blocked gas parameter

Relief rate, Case 1 ␮ Dynamic viscosity, Centipoises (Cp)

Relief rate, Case 2 ␷ Kinematic viscosity, m/s2
Temperature ⌬x Wall thickness, m
Pressure, Case 1 ␳ Density, kg/m3
Pressure, Case 2
Subscripts
i Inside
PR o Outside
ave Average
w Wall
N Normal operating condition
D Design condition
TN
R Relieving condition
PN

Time, t
Saeid Rahimi Mofrad is a process engineer with Petrofac
Gas is PSV
blocked in opens Engineering and Construction at Al Soor, Sharjah, UAE. He is inter-
ested in relief load calculation, overpressure protection systems
design and flare network sizing. Mr. Rahimi Mofrad has an MS
FIG. 2 Typical trend of blocked-gas parameters. degree in chemical engineering from the Sharif University of Tech-
nology and a BS degree from Shiraz University, Iran.

HIGH PRESSURE PUMPING SOLUTIONS

woodgroupsurfacepumps.com
54
I NOVEMBER 2010 HYDROCARBON PROCESSING Select 160 at www.HydrocarbonProcessing.com/RS
 PLANT SAFETY AND ENVIRONMENT SPECIALREPORT

Customize operator training

for your thermal oxidizers
This case history shows the benefits of site-specific programs
in new equipment installations
T. GILDER, Shintech Louisiana, LLC, Plaquemine, Louisiana; and
D. CAMPBELL, T. ROBERTSON and C. BAUKAL, John Zink Co. LLC, Tulsa, Oklahoma

S
hintech began construction of its new Plaquemine, Loui- For example, Fig. 3 illustrates the potential problem of blow-
siana, manufacturing facility (see Fig. 1) in October 2005. off if a burner is over-fired. Moving from left to right shows what
The facility is located on a 1,725-ac site, and it manufac- happens as the air/fuel mixture velocity is increased. The last
tures chlorine, caustic soda and vinyl chloride monomer (VCM). furnace, on the far right, depicts the danger of going beyond the
Historically, Shintech has manufactured only polyvinyl chloride design firing limit for the burner. The why also better prepares
(PVC). This new plant is Shintech’s first integrated complex. This operators to react to new situations that may not have been cov-
new manufacturing facility uses state-of-the-art environmental ered in formal training sessions.
technologies and is subject to the most stringent environmental The course content included basics that apply to any equip-
controls in the country. ment of this type, along with very detailed and specific informa-
Thermal oxidizers (TOs) are commonly used to treat volatile tion on the equipment in their particular installation. Materials
organic compounds (VOCs) and carbon monoxide (CO) emis- presented include:
sions because TOs have very high destruction and removal effi- 1. Combustion and thermal oxidizer basics
ciencies.1,2 Thermal oxidation can be defined as “the process of 2. Safety overview and warnings
oxidizing combustible materials by raising their temperature above 3. Overall equipment familiarization
the auto-ignition point in the presence of oxygen and maintaining
it at high temperature for sufficient time to complete combustion
to carbon dioxide and water.”3
Shintech partnered with the manufacturer of the TOs to offer
customized training to 37 of its plant operators for the start-up
of its new facility.a Fig. 2 shows part of the thermal oxidization
system installed at the Shintech facility. The TO training course
was offered with optional Continuing Education Unit (CEUs)
credits that were available to any students meeting the follow-
ing criteria: take (not pass) a pre-test, attend at least 80% of the
course contact time, pass (at least 80%) a post-test and complete
an anonymous course evaluation. The manufacturer also operates
a training organization that is accredited; the JZI b is authorized to
offer CEUs through its accreditation by the International Associa-
tion for Continuing Education and Training (IACET).

Course design. While the plant is responsible for all safety

practices and training, JZI provided training designed to give
operators at the new facility a good idea of both the what and why
of operating the TO system and associated equipment. All too
often, operators are trained, sometimes hurriedly and haphazardly,
by existing experienced operators. The new operators may learn
what to do, but not the why behind it. It is also fairly common for
long-time operators not to understand some of the basics because
they were never taught to them.
The why is important because it helps operators better understand
the cause and effect that can impact safety, thermal efficiency (and, FIG. 1 Shintech plant in Plaquemine, Louisiana.
therefore, operating costs), productivity and pollutant emissions.
HYDROCARBON PROCESSING NOVEMBER 2010
I 55
SPECIALREPORT PLANT SAFETY AND ENVIRONMENT
4. Detailed walkthrough of equipment
5. Detailed blower, boiler, absorber, scrubber and demister
details
6. Detailed walkthrough of P&IDs
7. Drawings review
8. Pre-startup and refractory cure out
9. Normal startup and shutdown
10. Logic demonstration and DCS screens
11. Normal maintenance
12. Troubleshooting
13. Drawings.
Each student received a three-ring binder containing the color
PowerPoint slides of the course. Adequate room was provided
for them to make notes in the manual as desired. Some students
received their manuals prior to the start of the class and came pre-
pared with questions to ask. Statistics for all types of training show
that retention of the material diminishes fairly quickly after the
training has been completed. The student manual can be quickly
and easily referenced as often as needed to refresh previously
learned information. Although operators do not generally receive
their own copy of the operation and maintenance manual, the
student manual contains much of the same information, includ-
ing many of the written operating procedures.
FIG. 2 Photo of part of the thermal oxidation system during
installation.
FIG. 3 Series of furnaces showing the progression toward
blow-off of a burner flame.
56
I NOVEMBER 2010 HYDROCARBON PROCESSING
Training. The training was conducted over three consecutive
days, followed by a fourth day about six weeks later on a couple
of specific pieces of the equipment. Although most of the time
was spent in the classroom, there were many short sessions
spent outside at the equipment to review and emphasize specif-
ics after reviewing the basics in the classroom. The plant had
not been started up yet, so the equipment was installed but not
operational. While this did not allow the operators to do live
training, it did permit operators from all shifts to attend classes
together during normal working hours. This produced signifi-
cant interaction and feedback between participants and with
the instructors. Another important aspect of the training was
that supervisors were present during most of the sessions, which
sent a strong message about the importance of the class.
The format of the training was designed to be very interac-
tive. While colorful PowerPoint slides (for example, see Fig. 4)
were used to guide the discussion, operators were encouraged
to ask questions and make comments at any time. This was
encouraged in part through subject-oriented fun games such
as poker and bingo. Every time a participant asked or answered
a question, they were given a random card from a poker deck.
For the poker game, the student with the best poker hand at
the end of the day received a prize. For the bingo game, cards
were drawn from a deck until someone had enough matching
cards to win. The more cards a student had, the more chances of
winning, so this encouraged continuous and frequent participa-
tion. Other token gifts were also given out during the training
as deemed appropriate by the instructors, for example, to a
student asking a particularly good question.
Short video clips and brief plant visits were used to break
up the lecture periods to help keep students engaged in the
materials. Videos are particularly powerful when demonstrat-
ing potential problems, such as flashback from a burner, that
may not have been previously experienced at a particular plant,
but which could happen under certain circumstances. This is
analogous to airline pilots who train in simulators to react to
situations they hope they never encounter, but for which they
are prepared to handle just in case.
The actual equipment drawings for this plant were used dur-
ing the training to help familiarize the operators with the equip-
ment and with the operating procedures. To make it even more
FIG. 4 Slide showing the 3Ts of combustion: time, temperature
and turbulence.
SPECIALREPORT PLANT SAFETY AND ENVIRONMENT

realistic, photos were taken of individual components that would
pop up on the drawings when clicked. For example, clicking on
the symbol of a valve would pop up a picture of the actual valve
in the plant as shown in Fig. 5. This personalized the training
and made it easier for the operators to connect the drawings to
the actual equipment.
Results. Identical 15-question pre-tests and post-tests were given
to the students to measure learning. The pre-test assessed students’
knowledge prior to taking the class. The average pre-test and post-
test scores were 52% and 99%, respectively. The difference between
the scores is an indicator of what was learned in the training.
Students were also given a questionnaire at the end of the course
to assess their level of satisfaction with the course. Students did not
put their names on the forms, although their names were checked
off a list to show they completed the evaluation, which is one of the
requirements for receiving credits for the course. A five-point Lik-
ert scale was used, where 1 = none, 2 = little, 3 = average, 4 = above
average, and 5 = great. Students rated each section of the course
according to their interest in the topic and its benefit to them.
There was also a space to write in any comments they may have
had on the topic. Fig. 6 shows the averaged results by interest and
benefit for each topic. The results show that, on average, students
found all topics to be of above-average interest and benefit.
Another part of the questionnaire asked students for written
comments on the instructors and material. Some of the instruc-

5
Interest Benefit

Rating
3

1
1 2 3 4 5 6 7 8 9 10 11 12 13
Topic #

FIG. 5 Animated P&ID with a picture of an actual control valve. FIG. 6 Student ratings of interest and benefit of each course
topic.

less waste
more profit
increased sustainability
When your steam trap population is managed correctly you can
reduce steam consumption and emissions, improve condensate return
and control of processes, while enjoying lower maintenance costs.

We are international steam specialists and many of our

customers worldwide are saving millions following a tailored
Spirax Sarco audit and implementation.

To join them:
Contact us via www.spiraxsarco.com/steamtrapsurvey to arrange
an initial survey and receive your FREE steam and condensate loop book.

Select 161 at www.HydrocarbonProcessing.com/RS

58
I NOVEMBER 2010 HYDROCARBON PROCESSING
 PLANT SAFETY AND ENVIRONMENT
tor comments included “Very Knowledgeable,” “Excellent” and
“Very Thorough.” Some students felt more time should have been
spent on startup, shutdown and troubleshooting, and less time
on the drawings. Content and coverage are always the challenge
with a group of students having a wide range of backgrounds and
experiences. Note: All students met the necessary requirements and
received CEUs for the class.
Outcome. Properly training plant operators is critical to ensure
that process equipment is operating safely, while maximizing effi-
ciency and productivity and minimizing pollution emissions.
Operators need to understand some basic information about the
equipment, as well as the details on their specific installation.
Although not always possible, it is particularly beneficial to have
all operators together in the same class to enhance discussion and
mutual learning. Training should be customized to the needs of
the plant and should incorporate techniques such as fun games to
promote interaction among the participants and instructors. Ide-
ally, there should be a “hands-on” portion of the training where
instructors use the actual equipment during demonstrations. Pre-
testing and post-testing are effective tools to show that operators
have learned the key points in the training. HP
NOTES
a John Zink Company, LLC (JZC) manufactures thermal oxidation systems
used to destroy unwanted wastes.1
b The John Zink Institute (JZI) is the training group for JZC and delivers
training both at its US headquarters in Tulsa, Oklahoma, and at customer
plant sites. JZI works with the plant to determine a suitable course agenda.
Introduces the new MisterMesh®
MAXCAP ™
mist eliminator with 43% more capacity!
EMERGENCY
SERVICE
800-231-0077
ͻ>ŽǁĞƌƉƌĞƐƐƵƌĞĚƌŽƉ
ͻ,ŝŐŚĞƐƚǀĂƉŽƌĐĂƉĂĐŝƚLJ
ͻ^ƵƉĞƌŝŽƌůŝƋƵŝĚĚƌĂŝŶĂŐĞ
ͻDŽƌĞĞĸĐŝĞŶƚΘĐŽƐƚĞīĞĐƟǀĞ
WĂƚĞŶƚƉĞŶĚŝŶŐĨŽƌƚŚĞ
^DĂdžĂƉ™ ŵŝƐƚĞůŝŵŝŶĂƚŽƌ͘
ͻhŶŝƋƵĞ'ĞŽŵĞƚƌŝĐĐŽŶƐƚƌƵĐƟŽŶ
Call today ĂŶĚůĞƚƵƐŚĞůƉĮŶĚƚŚĞďĞƐƚƐŽůƵƟŽŶ
ĨŽƌLJŽƵƌĚĞŵĂŶĚŝŶŐŵŝƐƚĞůŝŵŝŶĂƟŽŶŶĞĞĚƐ͊
Ask for an Independent Test Report
Select 162 at www.HydrocarbonProcessing.com/RS
SPECIALREPORT
LITERATURE CITED
1 Baukal, C. E., ed., The John Zink Combustion Handbook, CRC Press, Boca
Raton, Florida, 2001.
2 Schnelle, K. B. and C. A. Brown, “Thermal Oxidation for VOC Control,
Chapter 13,” Air Pollution and Control Technology Handbook, CRC Press,
Boca Raton, 2002.
3 Moretti, E. C., “Reduce VOC and HAP Emissions,” Chemical Engineering
Progress, Vol. 98, No. 6, pp. 30–40, June 2002.
Tim H. Gilder joined Shintech Louisiana, LLC, in 2006 as vinyl chloride monomer
production superintendent. Prior to joining Shintech, Mr. Gilder served in engineering
and supervisory positions at flexible polyurethane foam, furfural, polyvinyl chloride
and ethylene dichloride/vinyl chloride monomer production facilities. He earned a
bachelor’s degree in chemical engineering from the University of Mississippi.
Dale Campbell, P.E., is a senior design engineer at John Zink Company, LLC,
where he serves as the primary resource for incinerator troubleshooting and design
in the thermal oxidizer aftermarket group. Since 1973, his primary responsibility has
been the detailed design, equipment application, startup, and project management
of waste incinerator systems. Mr. Campbell earned a bachelor of science in chemical
engineering from the University of Tulsa.
Todd Robertson is a combustion service leader at John Zink Company, LLC. He is
responsible for thermal oxidizer installation supervision, startup, maintenance, service
and training. He retired from the United States Air Force after 23 years of service. Mr.
Robertson earned a bachelor’s degree from Embry Riddle Aeronautical University.
Chuck Baukal is the director of the John Zink Institute at John Zink Company,
LLC in Tulsa, Oklahoma. He has nearly 30 years experience in industrial combustion
in a wide range of industries. Dr. Baukal has a PhD in mechanical engineering from
the University of Pennsylvania and is a registered professional engineer in the state
of Pennsylvania. He has authored/edited eight books on industrial combustion and
has 11 US patents.
Select 163 at www.HydrocarbonProcessing.com/RS
59
 PLANT SAFETY AND ENVIRONMENT
Emergency response planning—
start at the plant design stage
Follow these guidelines for a safer facility
R. SAINI, Kuwait National Petroleum Company, Kuwait
A
n optimum level of emergency
preparedness can be achieved if
emergency response planning
is started early and the requirements are
incorporated in the plant design. Codes
and standards provide basic guidelines but
there is also a need to go beyond the books.
Emergency response and management
requirements should be established and
documented, and appropriate provisions
should be part of the system.
Incidents such as fires and explosions
continue to make news every now and
then. Such incidents, if not effectively con-
trolled, have serious impact on life, envi-
ronment, property and business. Adequate
and efficient emergency response can mini-
mize the consequences of the incidents,
should one occur. Emergency preparedness
therefore, continues to be an important
element of the overall process safety man-
agement system.
Fig. 1 shows the components of an
emergency preparedness system. It includes
the following:
• Emergency mitigation and control
systems that include process control and
emergency shutdown systems, and emer-
gency detection and alarm systems, passive
as well as active fire protection systems
• Emergency response and manage-
ment systems that include emergency plans
and procedures, communication systems,
logistics and resources.
Continual reviews, checking, mainte-
nance and testing of emergency prepared-
ness facilities are important to ensure
their effectiveness when needed. Overall
plant layout and design of the individual
facilities, utilities and support systems are
normally done based on operational and
maintenance requirements. These are done
in accordance with the applicable codes,
standards and regulations, safety guidelines
SPECIALREPORT
and insurance requirements. Emergency The emergency scenario register is the
planning activity should also be initiated at basic document for emergency prepared-
the project conceptual stage. Fig. 2 demon- ness planning. The data is also used for car-
strates how the emergency planning team rying further quantitative risk assessment
interacts with the plant design team. and consequence analysis.
Emergency preparedness requirements Fire protection philosophy document
must be established and documented. The should include the design basis for:
following documents should be developed • Firewater systems—firewater storage
as a part of the project emergency prepared- and pumping capacity, distribution net-
ness design package: work design and sizing, fire hydrant design
Emergency scenario register includes: and location
• Emergency scenarios (flammable/toxic • Fire protection requirements for
liquid/gas leaks, fires etc.) identified for each individual facilities such as process areas,
plant facility based on qualitative risk assess- storage tanks, operation control rooms,
ment for each plant facility, considering the electric substations, cooling towers, waste
products handled and operating parameters treatment areas, chemical storage and other
and carrying out credibility analysis. buildings
• Overall emergency scenario list for • Fire detection and alarm system basic
the plant along with the recommended requirements.
emergency response strategies for each type The philosophy document serves as a
of emergency. basis for detail design as well as for develop-
Emergency mitigation Emergency Emergency response and
and control systems preparedness management systems
Assessment and Testing and Training and
Plans and procedures
installation maintenance drills
Subsequent Emergency response
Communication Equipment and
Design stage periodic adequacy and management
system materials
reviews teams
Emergency coordination Assessment, Testing and
centers, fire station availability maintenance
FIG. 1 Emergency preparedness components.
HYDROCARBON PROCESSING NOVEMBER 2010
I 61
SPECIALREPORT PLANT SAFETY AND ENVIRONMENT

ing a fire protection manual for subsequent
use. The fire protection manual includes
details of the actual as-built fire protection
systems installed, along with testing and
maintenance requirements.
Pre-incident planning for a particular
plant facility/area includes:
• Listing all types of potential emer-
gency incidents and possible escalations
• Establishing an incident control strat-
egy and resource requirement, checking the
adequacy of the fixed emergency control
systems to control the incident
• Assessing external emergency response
requirements—manpower, equipment and
materials as necessary to supplement the
fixed systems and also as an alternate in case
of fixed-system failure
• Developing an emergency response
equipment deployment plan, including fire
hose laying requirements according to the
availability of fire hydrants.
The pre-incident plans should be devel-
oped for all plant facilities/areas. The worst-
case scenario requirements for each type of
emergency are used to verify the design of
the common emergency response facili-
ties. For example, pre-fire plans should be
used to review and update, if necessary, the
fire protection design philosophy and ade-
quacy of the fire protection and emergency
response systems, including the following:
• Firewater system design—design flow
and pressure, water storage, distribution
network layout and sizing, to meet the
worst-case fire scenario demand
• Review and update of the number
and location of fire hydrants
• Number and types of emergency-
response vehicles and materials.
Perimeter
fencing, facilities
arrangement,
safety distances
Codes, standards, regulations
The information from pre-incident
plans is also used for:
• Identifying gaps and possible
improvements
• Checking the adequacy of the plant
drainage system to handle firewater effluent
as well as any hazardous material discharge.
Emergency response and manage-
ment planning should also be started at
an early stage and the location and design
requirements for the following should be
established:
• Emergency communication systems
• A site emergency response center for
identified incidents such as fires, oil spills,
hazmat violations, chemical spills and per-
sonnel injuries
• A holding area for external response
teams
• Evacuation requirements
• An emergency management center—
emergency management team work sta-
tion.
Emergency communication require-
ments include:
• Automatic detection and alarm sys-
tems, manual call points
• An emergency communication con-
trol and emergency response team dis-
patcher center
• Personnel call-out communication
• Public address and siren systems.
Emergency evacuation requirements:
• Local assembly points—pre-identified
areas near each occupied building or work
area where the occupants gather in case of
an emergency. They wait for headcount and
further instructions.
• Main gathering areas and shelter-in-
place—pre-identified area in a safe loca-
Overall plant Emergency
Emergency response
layout response
plans
access
tion where the evacuees might stay for a
longer duration in case of a major plant
emergency. Such areas should provide pro-
tection against severe weather conditions
and be equipped with basic facilities such
as drinking water, toilet, first-aid and com-
munication. Large cafeteria dining halls,
auditoriums, basements, etc., if available,
may be considered for use as gathering
areas during emergency evacuations. Also,
consideration should be given to designing
buildings as a shelter-in-place for use in the
event of a toxic gas release emergency. It is
advisable to identify additional areas to be
used as alternate gathering areas.
• Evacuation routes—leading to local
assembly points and further to main gath-
ering areas to be identified and displayed
prominently at appropriate locations, along
with necessary instructions and guidelines.
Some important areas requiring special
consideration in the plant layout, fire water
system design, pre-fire planning and emer-
gency response planning are:
• Large hydrocarbon storage tanks
• Wind direction
• Plant drainage system.
Hydrocarbon storage tank fires have
always been a challenge for firefighters.
Tank fires have potential to escalate, if not
controlled in the initial stages, and the con-
sequences may be disastrous. This is more
so with large storage tanks and tanks con-
taining light products and products with
‘boil over’ potential, e.g. crude oil.
Special consideration should be given
while developing tank farm layout, espe-
cially when large tanks are part of the
design. Roads should be provided on at
least three sides of large-diameter tanks so
that sufficient access is available for fire
fighting appliances. Maximum cumulative
volume of the flammable liquid stored in
all the tanks in the dyke should be limited
to reduce risk. The Institute of Petroleum

Evacuation, (IP) Safety Code of Practice, Vol. 3 provides

Life safety, Facility design
Emergency planning

assembly Emergency guidelines in this regard.

passive fire points/shelters
protection, etc. Communication scenario Dykes or bunds are important aspects
systems identification
Emergency of designing tank farms. Codes permit
Process control coordination
systems Plant drainage dykes made of earthen, concrete or masonry
centers
Incident construction. Concrete or masonry dykes
Other systems Firewater Emergency control help save space and require low mainte-
systems communication strategies
nance as compared to earthen dykes. How-
Mobile ever, their use should be limited to low
Active fire
protection
fire-fighting Emergency heights (roughly 1 m) wherever possible.
equipment resources Pre-fire Concrete dykes restrict the ventilation and
systems –
(worst-case plans
philosophy/ dead-air pockets that are formed close to
Hydrant design, scenario)
manual
spacing the dyke base where flammable vapors can
accumulate. This is also a hindrance to fire-
FIG. 2 Emergency planning at the design stage improves emergency preparedness. fighters and it may be difficult for a person
to escape during an emergency. The dyke
62
I NOVEMBER 2010 HYDROCARBON PROCESSING
 PLANT SAFETY AND ENVIRONMENT SPECIALREPORT

layout dimensions can be worked out by plan needs to be prepared, considering the plan should include the estimated number
volume calculations, code requirements prevailing wind direction. This implies that of fire hoses along with the foam equipment
and accesses. Wherever possible, the fol- one side of the tank (downwind) will not requirements. Sufficient quantity of foam
lowing points should be considered: be available for placing foam equipment. concentrate should be made available.
• Maximize the tank spacing within the The wind direction changing during fire Wind direction is another important
same dyke. However, the tank spacing with fighting is still another constraint. Alternate factor that requires proper attention dur-
respect to the adjacent dykes should not be locations must be considered. An adequate ing plant design and emergency response
ignored. number of fire hydrants should be avail- planning. Drifting and pattern of a haz-
• Space between the dyke and the tank able at strategic locations to feed the foam ardous vapor cloud depends on the wind
shell should be such that the liquid jet from equipment. Fire hydrants should have out- direction. In case of a toxic gas release the
any potential leak from a hole in the tank lets to enable connecting of large-diameter occupied areas downwind of the leak source
shell is contained within the dyke. hoses for the purpose. Thus, the pre-fire will be affected. In case of a flammable gas
• Reach of the foam thrown from por-
table foam equipment is an important
consideration. Therefore, the road to tank
distance should be optimized, without
compromising other safety requirements,
to ensure effectiveness of the portable foam
equipment to be deployed as per the plan.
Pre-fire plans for storage tanks should
cover many scenarios including small leaks,
rim seal fires and the possible escalation to
full open-top fire. A fixed-foam application
and cooling systems attached to the tank
are the most effective means for controlling
tank fires. Floating roof tanks are normally
provided with fixed foam pourer systems
for rim seal fire protection only. However,
full-surface open-top tank fires, though
not very common, can and do occur if the
Full-Service
floating roof sinks.
In absence of the fixed foam systems
controlling full surface open top fires in the
Decontamination
large floating-roof tanks will require foam
application by using portable and mobile
equipment. Using portable and mobile Get the The right chemical is crucial
equipment may also be necessary when the expertise for process unit and
vessel decontamination.
fixed systems, where provided, might get that drives
damaged or become inoperative. There- results For optimal results – shorter turnaround times,
fore, it is necessary to ensure that adequate faster vessel entry, lower costs, improved safety –
provisions are made in the fire protection you need something more.
systems design. That’s why our patented Zyme-Flow® chemistries
For calculating the overall design fire- are backed by world-class service – including
water demand, water requirements for a comprehensive planning, chemical recommenda-
foam application over a full open-top sur- tions and project management. From the industry’s
face area, tank shell cooling for the tank on most experienced decontamination specialists.
fire and exposure protection of the adja-
If you’re ready to take decontamination results
cent tanks should be considered. It must be to the next level, we’re ready to help.
noted that the foam application rate is con-
siderably higher when foam is applied using
portable means as compared to the foam
application using fixed systems. Recent
research by specialist fire-fighting agencies
has recommended higher application rates Worldwide Leaders in Refinery
than those specified in the codes. Decontamination
High discharge foam monitors with Tel: +1 832.775.1600 Visit us at:
large-diameter hoses should be used for tank Toll Free: 877.332.6648 OSEA 2010
fire fighting. Deploying a large number of www.zymeflow.com Stand 4N4-06
the smaller-capacity foam monitors with
conventional 65-mm fire hoses may not be Zyme-Flow is a mark of United Laboratories International, LLC. ©2010 United
Laboratories International, LLC. All Rights Reserved
practical. Also, a foam monitor deployment
Select 164 at www.HydrocarbonProcessing.com/RS
63
SPECIALREPORT PLANT SAFETY AND ENVIRONMENT

leak, there are chances of the drifting flam-
mable vapor cloud coming in contact with
a source of ignition on the downwind side,
with potential danger of an explosion and
fire. A plant layout design team must con-
sider the prevailing wind direction, while
locating the potential critical sources of the
hazardous gaseous leak, with respect to the
populated areas/sources of ignition inside
as well as outside the plant boundaries.
It is important to understand the pre-
vailing wind direction concept. The aim
is to locate the potential toxic/flammable
gas leak sources in such a manner that the
chances of the vapor/gas cloud drifting
toward the ignition source or the populated
area are minimal. Some sites have a clear
single prevailing wind direction throughout
the year, while other sites may have dif-
ferent prevailing wind directions during
other times of the year. Therefore, designers
must study the wind rose diagrams care-
fully and arrive at a direction that has the
minimum chance on wind blowing in the
wrong direction. Such direction is often
termed as the cross-wind direction.
Drains and sewers also have an impact
on plant fire safety. Industry has witnessed
the consequences of the plan drainage sys-
tem designed without proper consideration
to fire safety, such as:
• Fire spreading through underground
oily sewers
• Explosion in sewers
• Fire spreading due to an oil layer on
the open drains
• Drains overflowing and water logging
during fire fighting.
Fire safety considerations related to
plant drainage systems are:
• Drainage systems must be designed
to handle the firewater discharge expected
during fire emergencies
• Oily sewers should incorporate liquid
seals to stop propagations of fire through
vapor spaces in the sewer
• Open drains, designed for non-oily
discharges in normal operations still have
chances of carrying oily effluent in abnormal
situations, and should also be provided with
liquid seals at certain intervals/locations.
Conclusion. Effective emergency
response is essential to minimize the con-
sequences of an incident in the petroleum
industry. Plant design must incorporate the
emergency response requirement, in addi-
tion to the normal plant safety requirements
laid out in the code and standard guide-
lines. Practical aspects of emergency plan-
ning and response should be incorporated
in the plant design, along with consulting
emergency response specialists. HP
Ramesh Chand Saini is a
Health Safety and Environment pro-
fessional with over 35 years of experi-
ence in emergency preparedness and
response systems in the hydrocar-
bon processing industry. He currently works at Kuwait
National Petroleum Company (KNPC) and his respon-
sibilities include developing and updating plans and
procedures for fire safety engineering and management,
emergency response and crisis management systems;
and also ensuring fire safety adequacy of the existing
plants as well as the new projects. Mr. Saini has initiated
a number of projects for upgrading fire safety facilities in
refineries. After graduating from the Indian Institute of
Technology (IIT) in Kanpur, India, he worked for 13 years
for Engineers India Limited (EIL) in its project engineering
team in the area of safety and fire protection for refiner-
ies and petrochemical projects. Mr. Saini is a member of
the Society of Fire Protection Engineers (SFPE), a certified
HAZOP leader and ISO 9001 lead auditor and has partici-
pated in a number of HAZOP studies and audits of fire
safety facilities at oil installations across Kuwait. He is a
member of a number of fire incident investigation teams.
Previously, Mr. Saini was a member of the fire committee
of the Bureau of Indian Standards, involved in develop-
ing a number of Indian standards on fire safety. He has
also been briefly associated with the Oil Industry Safety
Directorate of India for developing safety standards.

The Fundamentals of Piping Design

By Peter Smith
262 pages • Hardcover • Pub date: April 2007
ISBN: 978-1-933762-043 • Price: $175

Written for the piping engineer and designer in the field, this first
part of the two-part series helps to fill a void in piping literature,
since the Rip Weaver books of the ‘90s were taken out of print.

Advanced Piping Design

By Rutger Botermans and Peter Smith
250 pages • Hardcover • Pub date: May 2008
ISBN: 978-1-933762-18-0 • Price: $175

An intermediate-level handbook covering guidelines and

procedures on process plants and interconnecting piping systems.

The Planning Guide to Piping Design

By Richard Beale, Paul Bowers and Peter Smith
300 pages • Hardcover • Pub date: September 2010
ISBN: 978-1-933762-37-1 • Price: $175

The Planning Guide to Piping Design covers the entire process of

planning a plant model project from conceptual to mechanical
completion, and explains where the piping lead falls in the process
along with his roles and responsibilities.

To place an order, visit www.gulfpub.com or call +1 (713) 520-4426.

Select 165 at www.HydrocarbonProcessing.com/RS Select 166 at www.HydrocarbonProcessing.com/RS

64
 PLANT SAFETY AND ENVIRONMENT
Optimized fired heater control
Residual oxygen measurement principle
lowers emissions and improves efficiency
A. J. MOURIS, Hobré Instruments, Purmerend, The Netherlands
T
ighter emission regulations and high
energy costs pose new challenges to
control systems for fired heaters.
Rapid changes in fuel gas heating value,
air demand and composition are typical
for applications in oil refineries, chemical
plants and many other sites. Traditional
feedback control based on temperature,
stack oxygen and combustibles measure-
ment is not quick enough to handle rapid
changes effectively. This short coming is
typically addressed by controlling the excess
air set point with a certain safety margin.
Unfortunately, this approach prevents the
emission of unburned components while
increasing CO2 emission due to poor fuel
economy—air is heated unnecessarily and
heat transfer efficiency is reduced. NOx
formation is promoted as a result of higher
oxygen levels in the combustion process.
For these reasons, feed forward control of
the air/fuel ratio is gaining more attention.
Properly selecting and installing the fuel
gas property analyzer and using the right
control parameters are essential to get the
best results.
Control parameters. The control
system philosophy of fired heaters varies
depending on the requirements and heater
or boiler design. However, in all cases, the
furnace’s thermal load and the air/fuel ratio
are two critical parameters that must be
monitored and controlled.
Heat load control. Depending on the
control system design, the Wobbe Index
(WI), the heating value and gas density may
be required as input(s). The heating value is
the amount of heat produced when a unit
volume or fuel mass is burned stoichiomet-
rically. The higher heating value includes
the heat of water condensation formed in
the combustion process; the lower heat-
ing value does not. The WI (defined as the
heating value of a gas divided by the square
root of its specific gravity) is a measure of
the interchangeability of fuel gases when
introduced into a heater via a burner with
a fixed differential pressure. Two gases with
the same WI will deliver the same amount
of heat into a combustion process per unit
of time, regardless of the composition. To
clarify this concept, consider the following
fuel gas cases:
• Case 1—40% methane and 60%
hydrogen (by volume)
• Case 2—58% methane and 42%
nitrogen.
The lower heating value by volume
of these two gases is the same, i.e., 20.82
MJ/Nm3. The WI however, is 40.55 MJ/
Nm3 for Case 1 and 24.39 MJ/Nm3 for
Case 2! This means that the amount of
heat delivered per unit of time through
the same burner will be 40% lower in the
second case.
Air/fuel ratio control. The combus-
tion air flow supplied to an industrial
furnace is typically linked to the fuel gas
flow. In smaller installations, this may be a
mechanical link; in larger installations, air
and fuel gas temperature and pressure are
Flowmeter
Sample SG
Low
Pressure
reducer
High
Manifold
Heat
PI
exchanger
Air station Booster
FIG. 1 Typical residual oxygen content analyzer schematic.
SPECIALREPORT
taken into account. If large fluctuations in
the fuel gas composition are expected, the
signal from a WI analyzer or calorimeter is
used for correcting the air/fuel flow ratio.
Typically, the assumption is that there is a
proportional relationship between heating
value and air demand. Whereas this is cor-
rect for hydrocarbon-based fuel gases like
natural gas, for fuel gases containing sig-
nificant percentages of hydrogen, olefins,
CO2 and/or oxygen this approach fails.
The following are fuel gas cases:
• Case 1—100% hydrogen
• Case 2—88.5% methane and 11.5%
nitrogen.
The WI for both gases is the same,
i.e., 40.9 MJ/Nm3. The Combustion Air
Requirement Index (CARI) is defined as
the stoichiometric air demand divided by
the square root of the relative gas density.
Results: 9.0 for Case 1 and 10.9 for Case
2. This means that if the fuel gas com-
position changes from hydrogen poor to
hydrogen-rich composition the excess air
may be controlled 20% too high. Please
note that, instead of WI and CARI, a
similar case can be construed for heating
value and air demand; this follows from
the definitions:
SG Bypass
Bypass
Vent
ZRO2 cell
Restriction
Measuring oven
Mixing
chamber
Drain
HYDROCARBON PROCESSING NOVEMBER 2010
I 65
 PLANT SAFETY AND ENVIRONMENT SPECIALREPORT

WI = Heating value/√(Specific gravity)
CARI = Air demand/√(Specific gravity)
Residual oxygen content analy-
sis. In a typical residual oxygen content
analyzer sample, gas is continuously mixed
with combustion air under controlled
conditions followed by catalytic combus-
tion in an electrically heated furnace. The
residual oxygen content in the flue gas is
measured with an accurate and reliable zir-
conium oxide sensor. In the control unit,
the following combustion parameters are
calculated from the oxygen signal and the
(optional) density signal: WI, CARI, calo-
rific value (or BTU) and specific gravity.
• The sample handling system
should have minimal internal volume.
Although a fast-loop theoretically can
compensate for any dead volume in the
system, this will result in excessive vent-
ing and/or flaring of fuel gas. Ideally, the
analyzer has an integrated sample condi-
tioning system and requires no additional
external sample handling.
• A combustion air requirement
signal should be available. As discussed
previously, the heating value or WI can be
poor indicators for the air demand in fuel
gas applications. Therefore, it is important
to select an analyzer based on the residual
oxygen content principle that stores sepa-
rate calibration lines for WI/heating value
and CARI/air demand.
• The analyzer’s rangeability should
match all possible cases. Typically, the ana-
lyzer should be able to handle large fluctua-
tions in the fuel gas composition. Residual
oxygen content analyzers analyze fuel gases
of all possible compositions in the 0–120

OUR MISSION
The concept was first explored in the US,
but the European gas distribution compa-
nies Gaz de France and Dutch Gasunie have

YOUR SAFETY
really optimized the benefits. Their prime
objective was to develop an instrument that
was as fast as possible for optimizing natural
gas blending operations to meet grid entry
specifications. However, in the last decade
the technology has also proven to be very
suitable for fuel gas, vent gas, flare gas, bio-
GET MORE
WITH HART
gas and steel plant offgas applications.

Instrument installation and selec-

tion. When the decision is to install an MSA Ultima® X Series
analyzer for measuring the heating value Gas Monitors
and/or WI for feedforward fuel and air/fuel
ratio control, the following requirements
More efficient asset management
should be fulfilled:
More flexibility with digital
• The analyzer should be as fast as or analog capability
possible. It doesn’t make sense to install a
More compatibility with existing
calorimeter with a 20-sec response time if installed operations
changes occur within seconds. With residual
Ask about our new 10-year warranty
oxygen technology, a response time of less
on DuraSource™ Technology for Ultima
than 5 sec is achievable. XIR and XI Gas Monitors. For your gas
• Signal noise should be as low as detection solutions, contact MSA
possible. High signal noise levels require at 1.800.MSA.INST.
smoothing of the signal, typically by aver-
aging. As a consequence, the response
from the control system to a step-change
will be slower. A residual oxygen content To see videos about the
analyzer offers a repeatability of 0.05% of Ultima X Series of Gas
measured value. Monitors, scan with
• Local installation should be close your Web-enabled
mobile phone.
to the sample tap point. Ideally, the fuel * www.2dscan.com
gas heating value and air demand signal * standard data rates may apply

should be available before the fuel gas leaves

the burner tip. This means that the travel-
ling time of the fuel gas from the sample
tap point to the burner should be longer | G A S M O N I T O R S | S C B A | M U LT I G A S D E T E C T O R S |
than the traveling time from the sample H E A D / E Y E / FAC E P R OT E C T I O N |
tap point to the analyzer plus the analyzer
response time. Outdoor installation in 1.800.MSA.INST | www.MSAnet.com
hazardous areas is not a must, but it often
follows from the requirment of installation
close to the sample tap point.
HYDROCARBON PROCESSING NOVEMBER 2010
I 67
SPECIALREPORT
■ In a typical residual
oxygen content analyzer
sample, gas is continuously
mixed with combustion
air under controlled
conditions followed by
catalytic combustion in an
electrically heated furnace.
MODERNIZING
Level Measurement
Sump Level Solution
Sump level measurements and pump
control are often a maintenance
challenge. The VEGAPULS through-air
radar and VEGAFLEX guided wave
radar provide reliable continuous level
measurements. The VEGAMET 391
supplies up to six relays, offering
lead/lag pump control and giving
power to the measuring instrument.
www.ohmartvega.com
info@ohmartvega.com
800.FOR.LEVEL
Select 168 at www.HydrocarbonProcessing.com/RS
68
PLANT SAFETY AND ENVIRONMENT
MJ/Nm3 (0–3,000 BTU/SCF) range with-
out the risk for flame-out or overheating.
• Thorough application review. Each
application is different and proper review
is essential. Issues that should be consid-
ered include:
º Calibration gas selection—the “right”
calibrated gases give the best accuracy in
most cases, do not contain many compo-
nents (not more than two preferably) and
allow sufficient filling pressure even when
ambient temperature may be low.
80-90%
10-20%
º Parameters measured—besides
CARI and WI, specific gravity, heating
value and air demand may be required.
º Fuel gas hydrocarbon and/or water
dew point—it is not uncommon that fuel
gas is taken from a knock-out vessel. Care
must be taken that condensation does not
take place in sample lines or inside the ana-
lyzer. Ideally, all sample wetted parts should
be heated 10°C–20°C above the maximum
expected dew point.
º Sulfur content and presence of other
corrosive components— wrong material
selection can rapidly corrode and clog an
analyzer. Proper component selection and
analyzer design enables continuous opera-
tion even when more than 10% sulfur is
present.
º Overall response time—a lag time
analysis from the sample probe tip to the
analyzer signal output should be provided
to ensure compliance with the require-
ments. This is especially important when
high pressure gas lines must be analyzed.
º Ambient temperature range and haz-
ardous-area certification requirements—these
must be considered.
Conclusion. Feedforward control of
fired heaters utilizing fuel gas with varia-
tions in composition and properties may
result in considerable improved combus-
tion efficiency. Emissions may be reduced
and product quality and equipment life
time can be improved. Online analyz-
ers based on the residual oxygen content
method have the potential to deliver the
parameters to be measured within seconds.
However, a successful implementation of
the technology requires a clear understand-
ing of what is to be measured and how the
signals are used in the control loop. Also,
the location of installation and a properly
designed sample handling system are key
factors for success. HP
LITERATURE CITED
1 API RP 556, First Edition, May 1997.
2 Driedger, P., “Controlling fired heaters,”
Hydrocarbon Processing, April 1997.
3 Physical Properties of Natural Gases, N.V.
Nederlandse Gasunie, 1980. s
Albert Mouris is technical direc-
tor with Hobré Instruments BV. He
has more than 15 years of experience
in process analysis and sampling sys-
tems. Previously, Mr. Mouris worked as
an energy market analyst with the Dutch gas incumbent
Gasunie. He received a chemical engineering degree
from the University of Twente.
 PLANT SAFETY AND ENVIRONMENT SPECIALREPORT

Consider real-gas modeling

for turboexpanders
New visualization methods expose problems with traditional designs
K. KAUPERT, OC Turboexpanders, Irvine, California

E
very year turboexpanders generate millions of Euros in rev- Ideal gases and liquids. The simplest, most popular equa-
enue for hydrocarbon processing plants by removing heat tion of state for gases is the Ideal gas law; it states:
from gas streams, also known as the “turboexpander refrig- P = ␳RT (1)
eration benefit.” To maximize this financial benefit, accurate gas For liquids, the bulk modulus and coefficient of thermal expan-
dynamic performance predictions for turboexpanders are a neces- sion combine for the equation of state given as:
sity. This requires an accurate thermodynamic equation of state
that uses a real-gas model. But, which real-gas model is best? For
example, an ideal gas assumption can cause horribly wrong perfor-
mance predictions due to gas compressibility at high pressures and
low temperatures.1 As a result, all turboexpander manufacturers use
real-gas models in their simple gas dynamic sizing predictions.
However, beyond the simple sizing predictions, advanced tur-
boexpander manufacturers continue to apply real-gas modeling
in the detailed computational fluid dynamics (CFD) design of its
turboexpanders. This permits turboexpanders, such as shown in
Fig. 1 to attain high efficiency levels. Historically, the application
of real-gas models in commercially available CFD packages has
been problematic or even non-existent. This is due to:
1. Most turbomachinery manufacturers (e.g., those producing
gas turbines or turbochargers) are content with an ideal gas law or
a simple real-gas model since their compressibility effect is modest
and the flow is single phase. FIG. 1 Turboexpander designed with real-gas modeling in CFD.
2. Increased computational time is required when a real-gas
model is applied in CFD (slower code).
3. Numerical robustness is decreased when a real-gas model is
applied in CFD (code can more easily diverge).
4. Difficulty modeling two-phase flow in a wet gas expansion
region.
The turboexpander manufacturer is, therefore, confronted with
somewhat unique challenges, as the expander inlet gas can exhibit
substantial compressibility while the expander outlet gas can exhibit
two-phase wet gas flow.2 For example, Fig. 2 shows an expander
impeller connected to a rotating assembly. This particular impeller is
subjected to both compressible gas and wet gas, which requires accu-
rate real-gas modeling in CFD to maximize expander efficiency. But
there are many real gas models available in the open literature. So,
the question again arises, which model is best? In this article, we will
assess several real-gas models used in CFD for turboexpanders.

CONSIDERING THE EQUATION OF STATE

An equation of state relates a fluid’s state variables.3 For turboex-
panders, the equation of state is a thermodynamic equation, which
mathematically describes the interaction of the macroscopically
measurable properties of the process fluid (e.g., the thermodynamic FIG. 2 Expander side of a turboexpander rotating assembly.
variables of pressure, temperature, density and composition).
HYDROCARBON PROCESSING NOVEMBER 2010
I 69
SPECIALREPORT PLANT SAFETY AND ENVIRONMENT

Kl = dP/(d␳/␳) and ␣l =(1/V)dV/dT (2) ematics of these equations of state are presented elsewhere; they
Over limited ranges of pressure and temperature and without are large in size and require detailed explanation, among them
phase changes, these two equations ( Eqs. 1 and 2) give reasonable Refs. 4 and 5. The American Gas Association (AGA-8) is not
property predictions (e.g., the ideal gas law for ethane at less than 5 evaluated here as it poorly predicts the weight fraction of liquid
bar pressure and higher than 100°C temperature gives less than 2% in the wet-gas region at the expander outlet.
error on the density). But turboexpanders routinely handle gases Another very common and simple real gas equation of state
outside the limited range of the Ideal gas law and with two-phase utilizes the compressibility factor, Z and is called the universal
flow. In essence, real gas and liquid modeling is obligatory. gas law, given as:
P = Z␳RT (3)
Real gases and liquids. The need for accurate equations of Unfortunately, the determination of Z for gas mixtures in
state has resulted in an abundance of real-gas models in the open turboexpanders is very difficult; this equation (containing Z ) is
literature. So numerous are the real-gas models that they could not not evaluated here.
possibly all be evaluated in this short article. However, these five
real gas equations of state do find widespread application: CFD WITH REAL-GAS EQUATIONS OF STATE
1. Peng-Robinson (PR) CFD using a real-gas equation of state is a time-consuming
2. Soave-Redlich-Kwong (SRK) task. In the past, this author has pursued three methods to incor-
3. Benedict-Webb-Ruben-Starling (BWRS) porate real-gas equations in CFD. The first method is to program
4. Lee-Kesler-Plöcker (LKP) the real-gas equation of state and patch it into the CFD code.
5. GERG. While this method is the most straightforward, it frequently leads
These five equations of state are compared here with CFD to numerical convergence difficulties in the transition region
results through the expander side of a turboexpander. The math- from single-phase gas to two-phase wet gas. This concern can be
addressed with numerical damping routines that are ingeniously
applied by some CFD vendors. The second method is to use the
real-gas equation of state to prepare a set of “look-up” tables in the
desired pressure and temperature range of the gas. This method
is more tedious for pre-processing the CFD but tends to avoid
convergence problems while running the code. A third method
to follow uses the tables from the second method to create a set of
polynomials to approximate the equation of state. But the third
method can lead to large errors outside a prescribed temperature
and pressure range and convergence difficulties for the CFD.
TABLE 1. Gas conditions for the expander side of the
NGL hydrocarbon example
Gas dynamics
Job: Example NGL fractionation
Component MW Expander, mol%
Methane 16.0430 81.000%
Ethane 30.0700 11.000%
Propane 44.0970 5.000%
FIG. 3 Density as a function of pressure and temperature from a i-Butane 58.1230 0.800%
real-gas model and used in a “look-up” table for CFD.6
n-Butane 58.1230 0.900%
i-Pentane 72.1500 0.344%
275 K n-Pentane 72.1500 0.100%
7.0 MPa 5.5 255 n-Hexane 86.1770 0.050%
6.0
6.5 5.0 4.5 4.0
3.5 MPa 245 C7+ 110.00 0.002%
Nitrogen 28.0134 0.800%
235
Carbon dioxide 44.0100 0.003%
a Water 18.0153 0.001%
Enthalpy

225 K
b
Total 100%
Given process conditions Rated case expander
d Molecular weight 20.087
c Inlet pressure, P1, MPa a 6.91
Inlet temperature, T1, °C 0.2
Outlet pressure, P2, MPa a 3.48
Entropy
Mass flow, kg/sec 13
FIG. 4 Enthalpy vs. entropy diagram for a hydrocarbon gas
example. Volume flow, Nm3/h 52036

70
I NOVEMBER 2010 HYDROCARBON PROCESSING
 PLANT SAFETY AND ENVIRONMENT SPECIALREPORT

Most CFD vendors have followed the direction of the second surements for similar gas compositions and conditions, the GERG
method or at least opened up their code for input from such “look- model was found to be most accurate for this gas case. Although,
up” tables. It is the second method that the author has applied the LKP and BWRS models would be acceptable as well.
here. Essentially, the real-gas equation of state is used to generate Results from applying the GERG model through the use of
the thermodynamic “look up” tables for P=P (u,␳), T =T(u,␳), P generated property tables in the CFD are seen in the values given
= P (h,s), ␳=␳ (h,s), s=s(h,P), h=h(s,P), u=u (P,T) and ␳=␳ (P,T) in Figs. 5 and 6 at the expander design point. Fig. 5 shows the
where the variables are: distribution of temperature as predicted with the GERG real-gas
P Pressure equation of state in the natural gas mixture. A rapid temperature
T Temperature change is seen through the expander nozzles; it is typical for turbo
␳ Density expanders, as the temperature decreases due to the acceleration of
u Internal energy the flow. In Fig. 6, the relative velocity vectors in the expander
h Enthalpy impeller are seen, again as computed with the GERG real-gas
s Entropy. model. The overall image is shown on the left side of the figure
An example of such a “look-up” table is plotted in Fig. 3, taken and a zoom at the expander impeller trailing edge is seen at the
from Ref. 6.

Example: CFD for real gases in natural gas liquids.

To protect client confidentiality, a generic gas composition and
condition were selected for this example. However, both gas
composition and condition are representative of a natural gas
liquids feedstock to the expander side of a turboexpander. The
gas composition and condition are seen in Table 1. The enthalpy
vs. entropy diagram is seen in Fig. 4 along three lines for the gas
expansion. The red line “a-b” represents an isenthalpic gas expan-
sion as would be experienced through a Joule-Thompson valve.
The blue line “a-c” represents an isentropic, or perfect, expan-
sion of the gas. The green line “a-d” represents the gas expansion
through the turboexpander and indicates a lower outlet tempera-
ture than the isenthalpic red line expansion.
This highlights the benefit of using a turboexpander vs. a Joule-
Thompson valve for the gas expansion—a colder gas outlet tem-
perature with the green line and heat removal from the expander
gas stream. Table 2 shows the outlet gas conditions at the point “d”
compared between the ideal gas and five real gas models. It is seen FIG. 5 CFD results for the temperature reduction in the expander
immediately that the ideal gas model gives the lowest expander per GERG real-gas model.
outlet temperature along with the largest
expander wheel (impeller) output power.
However, the compressibility Z is seen as
1.0, which is not realistic for the expander
inlet or outlet gas conditions. Accordingly,
applying an ideal gas model would lead to a
substantial modeling error for this gas.
Among the five real-gas models, the PR
model gives the lowest expander outlet tem-
perature, lowest output power, and lowest
efficiency. The GERG, LKP and BWRS
models all give similar outlet temperature, FIG. 6 Left: CFD results using the GERG real-gas model for the relative velocity vectors in
the expander impeller (wheel). Three blades are shown along the shroud showing the
output power and efficiency, as well as, simi- parallel flow to the blade at the outlet. A zoom is shown to the right at the impeller
lar weight liquid percentages. Based on mea- outlet.

TABLE 2. Comparison table of CFD results from five real gas models at the expander outlet
Ideal-gas model Real-gas model Real-gas model Real-gas model Real-gas model Real-gas model
Computed outlet conditions Ideal GERG PR LKP SRK BWRS
Outlet temperature, T2, °C –33.7 –28.8 –29.4 –28.9 –28.3 –28.7
Zin/Zout 1.0/1.0 0.72/0.78 0.69/0.76 0.71/0.79 0.70/0.77 0.71/0.78
Specific enthalpy ⌬Hs, kJ/kg 86.84 54.03 52.12 54.37 52.52 53.69
Expander isentiopic efficiency, % 82.7 85.3 84.6 85.4 84.8 85.2
Expander impeller power, kW 933.2 599.1 573.2 603.6 579.0 594.7
Outlet weight liquid, % unable 13.60 12.74 13.52 13.32 13.64

HYDROCARBON PROCESSING NOVEMBER 2010

I 71
SPECIALREPORT PLANT SAFETY AND ENVIRONMENT

TABLE 3. Generalized recommendations regarding Finally, interested readers are always encouraged to review
applicable equations of state comprehensive texts on turbomachines, (such as Ref. 7) or more
elementary books on turboexpanders (Ref. 8).
Gas type GERG PR LKP SRK BWRS
As energy conservation has become one of the world’s foremost
Hydrocarbon +/– 2% +/– 5% +/– 2% +/– 3% +/– 2% priorities, the importance of efficient turboexpanders continues
Air +/– 2% +/– 5% +/– 3% +/– 3% +/– 2% to increase. HP
CO2 +/– 3% +/– 6% +/– 4% +/– 4% +/– 1%
H2 +/– 4% +/– 7% +/– 4% +/– 2% +/– 5% LITERATURE CITED
1 Beinecke, D. and K. Lüdtke, Die Auslegung von Turboverdichtern unter
NH3 not good not good not good not good not good Berücksichtigung des realen Gasverhaltens, VDI-Berichte 487, VDI-Verlag
Note: the PR equation of state was the least accurate for all gas types but it is also the sim- Düsseldorf, pp. 271–279, (in German), 1983.
plest to apply. For ammonia, none of the five real-gas models performed satisfactorily and 2
in-house models are still relied upon. Kaupert, K. A., “Design of Two-Phase Flow Air Separation Turboexpanders,”
Cryogenic Technology Journal China, Vol. 1, pp. 47–52 (in Chinese), 2010.
3 Zemansky, M. W., Heat and Thermodynamics: An Intermediate Textbook,
right, showing that the relative velocity vectors closely follow the
McGraw-Hill Book Co., 6th edition, 1981.
blade surface without any recirculation zones. Such visualizations 4 Modisette, J. L., “Equation of State Tutorial,” Pipeline Simulation Group
are the key to optimizing the gas path and efficiency of expanders (PSIG), Paper 0008-2000, 2000.
by reducing any unwanted entropy generation. 5 Kunz, O., R. Klimeck, W. Wagner and M. Jaeschke, The GERG-2004 Wide-
Range Equation of State for Natural Gases and Other Mixtures, VDI-
Berichte 557 Reihe 6, VDI-Verlag Düsseldorf, 2007.
Real-world view. Real-gas modeling is needed for accurate 6 Numeca, Numeca Fine Users Manual Version 6.1-1, Numeca International,
gas dynamic performance predictions in turboexpander CFD to February 2003.
optimize expander efficiency. As the five models have shown, dif- 7 Lüdtke, K., Process Centrifugal Compressors, Springer Verlag, Heidelberg,
ferent results are attained by applying different real-gas models. It 2004.
8 Bloch, H. P. and C. Soares, Turboexpanders and Process Applications, Gulf
is important for turboexpander manufacturers to use test results Publishing Co., Houston, ISBN 0-88415-509-9, 2001.
and to also obtain detailed field feedback. Together, this will allow
selecting the real-gas model best suited to a particular application.
In this article, only one example was presented. It is not wise to
generalize on the basis of just this one example. The scope is Dr. Kevin Kaupert is the director of technology at OC Tur-
boexpanders. He holds a doctorate in turbomachinery engineering
widened with other published or in-house data. Table 3 offers a from the ETH Zurich Swiss Federal Institute of Technology. He has
number of generalized recommendations as to which equation of over 25 years of experience in turbomachinery for cryogenics,
state should be considered for CFD in modern turboexpanders. power generation and aerospace applications.

presents . . .

Wednesday, November 10, 2010 11 a.m. ET / 10 a.m. CT

LIVE WEBCAST
“Run your pumps like a Pro: Tips for Boosting
Production and Reducing Risk at the Refinery”
With the myriad of engineering, logistical and safety challenges involved in a refining operation, common API pumps can often be
overlooked—as a potential source of improved productivity, or a cause of catastrophic failure if not operated properly.
In this webcast, Dan Kernan and Eddie Choung of ITT’s Industrial Process business will share best-practice advice to help attendees improve
the effectiveness of the pumps they use in oil and gas processing.

t
0QFSBUJPOBM
EPT
BOE
EPOUT
GPS
VTJOH
QVNQT
QSPQFSMZ
JO
SFmOJOH
BQQMJDBUJPOT

t
"O
PWFSWJFX
PG
DPOEJUJPO
NPOJUPSJOH
PQUJPOT
UP
JNQSPWF
TBGFUZ
BOE
SFEVDF
NBJOUFOBODF
DPTUT

t
5IF
TUBUVT
PG
"1*

DIBQUFS

GPS
QVNQ
SFQBJST

t
$BTF
TUVEJFT
UIBU
JMMVTUSBUF
UIF
SFBMXPSME
JNQBDU
PG
EJõFSFOU
BQQSPBDIFT
UP
QVNQ
PQFSBUJPO
BOE
NPOJUPSJOH
This webcast will provide practical advice for anyone who oversees the use of pumps—including refinery managers, maintenance/reliability
engineers and production supervisors.
ITT presenters:
Dan Kernan, Manager–Monitoring and Control
Eddie Choung, Aftermarket Sales Manager—PRO Services North America

REGISTER at
www.HydrocarbonProcessing.com
Select 169 at www.HydrocarbonProcessing.com/RS
 PLANT SAFETY AND ENVIRONMENT
Predictive emissions monitoring
helps reduce stack air emissions
New technology reduces compliance costs while optimizing operations
R. HOVAN, Rockwell Software Environmental Solutions,
Rockwell Automation, Austin, Texas
T
he oil and gas industry is under unprecedented environ-
mental scrutiny. Already struggling to cope with increas-
ingly rigorous air quality regulations, the industry’s been
battered in recent headlines about oil spills from the Gulf of
Mexico to Michigan’s Kalamazoo River.
To protect both the environment and company reputations,
producers must do far more than prevent potential environmental
crises. They need to employ solutions and invest in technology to
make oil and gas production cleaner. Today, oil and gas processers
across the globe are facing stringent emissions monitoring, report-
ing and reduction goals from national regulatory bodies, as well
as an ever-growing patchwork of state and regional regulations.
Carbon dioxide (CO2)—the primary greenhouse gas (GHG)—
and nitrous oxide (N2O)—a GHG largely blamed for acid rain
and ozone depletion—are among the top targets for reduction.
Traditional emissions monitoring and management relies heav-
ily on hardware, manpower and a variety of reporting forms
and spreadsheets. Today’s and tomorrow’s regulations require a
level of accuracy, timeliness, consistency and security that legacy-
emissions monitoring systems cannot easily provide.
The good news is that cost-effective, software-based predic-
tive-emissions monitoring systems (PEMS) and environmental
management applications already are available to help reduce
emissions. These systems use existing plant process-monitoring
equipment to measure emissions factors in real time, predict
future emissions scenarios and compile emissions reports, thus
helping to reduce the cost of regulatory compliance.
Any business manager who’s well-versed in efficient produc-
tion will recognize the potential uses for such accurate, immediate
data collection and analysis. Oil and gas producers already facing
increasing capital and operational costs can use this data for more
than achieving regulatory compliance. It can also help reduce
energy usage and improve productivity and business performance.
The technology for energy efficiency that actually improves com-
petiveness is here and it’s proven.
Regulatory ramp-up. Petrochemical and natural gas plants
face new emissions-related regulations that require immediate
action. New rules aim to prepare the industry for emissions reduc-
tion by focusing on monitoring and reporting improvements.
Plants that produce combined emissions (from combustion, flares,
fugitive emissions and vents) equal to or greater than 25,000 met-
ric tons of CO2equivalent (CO2e) annually must monitor emis-
SPECIALREPORT
sions and provide comprehensive reporting on an annual basis.
This is according to the 2009 GHG Mandatory Reporting Rule
from the US Environmental Protection Agency (EPA).
Simply put, the new regulations require that producers know
and report precisely what is coming out of their smokestacks. The
rule requires monitoring systems to be in place by Jan. 1, 2011.
The first annual emissions report is due March 31, 2012. After
enacting the 2009 rule, the EPA recognized that implementing a
traditional monitoring system might be cost-inhibitive for some
producers. As a result, the EPA adjusted a number of the require-
ments outlined in the initial rule to help reduce the burden.
Under the updated rule, facilities may measure emissions through
engineering estimates and emission-modeling software.
Despite these changes, oil and gas producers have encountered
many challenges in trying to comply with requirements. Process-
ing is a complex operation and emissions can vary widely, depend-
ing on a spectrum of variables ranging from the composition of
the fuel feedstock to weather conditions.
Many producers today are finding that they don’t know if
they’ve surpassed the regulated limit for specific emissions until
they prepare their required reports after the fact. This lack of real-
time monitoring makes it all too easy to surpass emissions limits.
So much so, that some producers even build noncompliance
charges into their business models.
Even producers with some emissions monitoring in place lack
complete knowledge about how much they are emitting and when
levels peak. Not only does this make compliance difficult, but it’s
nearly impossible to take emissions output into account in opera-
tions and business decisions.
Once oil and gas producers and the EPA have a clear grasp of
actual emissions levels, the next step in the evolution of emissions
regulation will be control. Everyone in the industry is watching
the proposed Transport Rule, which will replace the Clean Air
Interstate Rule (CAIR). The new law will set limits on emissions
from utilities and lay out cap-and-trade requirements. This could
foreshadow what is in store for the oil and gas industry.
By anticipating and preparing for upcoming regulations now,
producers can eliminate waste and make improvements to their
bottom line. The ability to closely track their emissions, and tie
this output information back to data based on process inputs,
opens up a vast assortment of opportunities to use this informa-
tion to reduce energy consumption and optimize production for
more efficient operations.
HYDROCARBON PROCESSING NOVEMBER 2010
I 73
SPECIALREPORT PLANT SAFETY AND ENVIRONMENT

A Texas-based petrochemical producer, driven by state emis-
sions reduction requirements, implemented a PEMS to monitor
and verify the continued reduction of NOx emissions by 50%. As
a result, the company began to operate well within permitted NOx
levels and is now able to sell (vs. buy) NOx credits through the
Texas Commission on Environmental Quality Emissions Banks
and Trading program.
The PEMS provided detailed information on boiler output
vs. input allowing the company to run its energy center’s four
300-MW boilers at their lowest oxygen levels by eliminating their
artificial oxygen floor. The plant improved energy efficiency by
1%, according to a Department of Energy calculation based on
temperature in the stack. Through its energy-efficiency savings
alone, the plant saw full payback on the project in 13 months.
CO
NOx
Efficiency
Fig. 1 illustrates the formation of NOx and CO as a function
of excess O2 within a typical combustion application. Efficiency is
also shown as a function of excess O2. As excess O2 decreases, NOx
decreases while CO and efficiency increase. To minimize NOx pro-
duction and maximize efficiency, excess O2 needs to be minimized;
however, equipment safety limitations and environmental regula-
tions limit the amount of CO that can be produced. As a result,
excess O2 can only be reduced until the CO constraint is reached.
At this point, combustion has been optimized, with the unit operat-
ing at the maximum allowable CO production, which corresponds
to the minimum NOx production and maximum efficiency.
Continuous emissions monitoring (CEM) systems.
New regulations will be the impetus for many oil and gas pro-
cessors to invest in a continuous emissions monitoring system.
However, when choosing which system to implement, processors
should carefully consider a system that not only eases compliance,
but that can also help improve productivity and reduce waste at
the same time. The primary distinction between systems on the
market is hardware- vs. software-based technologies.

Hardware-based CEM systems. Hardware-based CEM

Maximum systems require a significant investment in new equipment. The
efficiency
NOx, efficiency

systems require probes to be fitted to the wall of a plant’s smoke-

stacks to take samples of process gas at a continuous basis. These
CO

probes require weatherproof enclosures to protect them from the

extreme conditions of the stack. All equipment with exposure to
process gas must be designed and built for operation in hazard-
ous locations. Additional cabinetry and shelters are required for
the sampling equipment, gas analyzers and the control system
that will feed emissions information back into a computer for the
CO limit Minimum NOx EPA-required data collection and reporting.
The two primary types of hardware-based CEM systems are
direct extractive and dilution extractive. In a direct extractive sys-
0.00 1.00 2.00 3.00 4.00 5.00
O2, %
tem, a sample pump located in a shelter fitted to a smokes stack
pulls a sample of emissions through a heated probe. The probe
FIG. 1 NOx and CO efficiency. provides filtering via a 2-μ ceramic element. The hot/wet sample
continues through a heated sample line and onto a sample chiller
located within the CEM cabinet. The sam-
ple is then flash-cooled to 4°C and water is
Direct extractive CEM system rapidly removed. The now dry and very cold
Stack wall Analyzer sample is delivered to the analyzers on a dry
Undiluted Undiluted
hot wet dry sample basis. The analyzers, in turn, provide an out-
Extraction sample Cooler Filter and put to either a datalogger or programmable
PLC/DL
probe dryer flow control logic controller (PLC). A direct-attached
Heated probe Undiluted storage (DAS) polling computer collects this
and sample line dry sample data and produces reports for the EPA and/
Analyzer DAS
Hot wet (option) or state compliance.
Analyzer In a dilution extractive CEM system,
Dilution extractive CEM system the sample is immediately mixed with or
Stack wall Diluted Analyzer “diluted” by clean, dry and extremely cold
Diluted wet sample
air (at –40°C) within the probe. Approxi-
Dilution wet sample Pneumatic mately 5L of this air is sent up an umbilical
probe PLC/DL into the probe. This air is used to develop a
controller
Diluted
vacuum on a critical orifice located within
wet sample the probe. This orifice allows a precise
Analyzer DAS
amount of process gas to blend with the
Analyzer clean dry air delivered to the probe. This
diluted process gas flows down the sample
FIG. 2 Hardware-based CEM systems. tube in the umbilical under positive pres-
sure. In most applications, it is not necessary
74
I NOVEMBER 2010 HYDROCARBON PROCESSING
 GULF PUBLISHING COMPANY EVENTS
20
11
SAVE THE DATE
PROCESS CONTROLS AND INSTRUMENTATION CONFERENCE
9-11 March 2011 • Moody Gardens, Galveston, Texas
Hosted by both World Oil and Hydrocarbon Processing, the Process Controls & Instrumentation Conference will be
devoted to advancing process control and instrumentation in the oil and gas industry. www.GulfPub.com/PCI

INTERNATIONAL REFINING AND PETROCHEMICALS CONFERENCE

June 2011 • Beijing, China
The International Refining & Petrochemicals Conference is a market-leading program for technical and operating
management in the Hydrocarbon Processing Industry. This conference will offer you an effective means to market
to engineering and operations management in the HPI. Like Hydrocarbon Processing, the International Refining
Conference focuses on providing the industry the very best technical content. www.GulfPub.com/IRPC

MARKETING IN THE OILFIELD CONFERENCE

August 2011 • Houston, Texas
The Marketing in the Oilfield Conference provides an environment to learn new ideas and strategies in addition to
numerous opportunities to network with fellow upstream and downstream marketing peers. This conference focuses
on industry hot topics related to marketing, social media, communication issues and includes featured keynote experts
and presentations relevant to the topic in focus. www.GulfPub.com/MITO

WOMEN’S GLOBAL LEADERSHIP CONFERENCE

IN ENERGY & TECHNOLOGY
October 2011 • Houston, Texas
Hosted by both World Oil and Hydrocarbon Processing, the Women’s Global Leadership Conference in Energy &
Technology is the largest women’s event in the industry, and the only one that focuses on discussing the industry’s
key environmental, economic, professional development and human capital issues in one setting. Attendees leave the
conference with an increased understanding of the full range of pertinent issues and an increased ability to be change
agents in our industry. This conference continues to encourage the growth and leadership of women in the industry
and the respect and knowledge of energy and technology. www.WGLNetwork.com

For more information about Gulf Publishing Company events or to work with us to create a new event,
visit www.GulfPub.com/Events, e-mail Events@GulfPub.com, or call +1 (713) 520-4475.

(Event topics and dates are subject to change.)

GULF
P U B L I S H I N G C O M PA N Y
SPECIALREPORT PLANT SAFETY AND ENVIRONMENT

Views Real-time environmental

management reports
(optional)

Metadata warehouse
(optional) Browser-based
Control console
client

Sensor validation Ti No
Tx´ Model analytic engine
Predictions Dif. Greater than
Tx´, Fy´, ... Pz´ Inputs

Application server
Cache
Tx tolerance Outputs
Fuel flows
Txv, Fyv, ... Pzv Fuel quality NOx
Yes O2
Air flow
Validated sensor CO
Sensor models Alarms Process O2
data Temperatures

Cache
Ambient Emission alarms
Ti´ Humidity Sensor alarms
Tx, Fy, ... Pz Data reconciliation
Raw values
Data qualification and
sensor validation

Manual
DCS PLC UMS Historians Analyzer
entry

FIG. 3 Predictive emissions monitoring system.

to heat the sample line. In other applications, the sample dew
point hasn’t been sufficiently lowered, and heating is required.
The process gas is now delivered to the analyzers on a wet basis.
No water was removed from the sample. It was only diluted. The
analyzers then provide an output by the same method as direct
extraction systems.
Fig. 2 illustrates that the hardware-based approaches rely on
precise instrumentation and sensors. They require high installa-
tion costs, real-estate requirements and sustained skilled main-
tenance. In continuous emissions monitoring and reporting,
even small human errors can lead to inaccurate reporting, missed
deadlines, permit violations, financial penalties and ultimately lost
profit. Hardware-based approaches lack the accuracy, timeliness,
consistency and security demanded by today’s regulations.
Software-based CEM systems. A software-based CEM,
known as a PEMS, can help oil and gas producers cost-effectively
monitor emissions data without installing and maintaining expen-
sive hardware and systems.
The EPA Emissions Measurement Center has recognized
predictive monitoring as a viable alternative to hardware CEM
systems since 2005 and most states have followed the EPA’s guid-
ance.* Predictive monitoring continuously monitors emissions by
developing an online model using historical and real-time data
from existing plant sensors. This hybrid modeling technology
incorporates nonlinear empirical models such as neural networks,
*TheEPA’s Emission Measurement Center developed and published a PEMS
protocol and performance specifications, available on the EPA’s TTN
website—http://www.epa.gov/ttn/emc/perfspec/ps-16.pdf
76
I NOVEMBER 2010 HYDROCARBON PROCESSING
as well as first-principles models to provide the most accurate
prediction models available in the industry. These models, crucial
to accurate monitoring and compliance, can also be used to better
understand and, therefore, control energy use.
Models are executed online, using a proprietary analytic engine
to provide real-time predictions of emissions from a wide range of
sources and fuels. Model validation is a routine that applies known
values to the sensor inputs and verifies the values against known
outputs. Predetermined input values are applied to the PEMS and
output values are then calculated. These values are compared to
the known output values from known input values developed dur-
ing modeling and relative accuracy test audits (RATAs). Values are
compared and the software determines that they are in accuracy
compliance. This process is an equivalency to an EPA-mandated
quarterly audit.
Achieving the EPA-mandated minimum operational standard
for demonstrating continuous compliance with PEMS requires
the ability to continue to provide reliable data in the event of
sensor failure. This requires a methodology for detecting and com-
pensating for those failures. The predictive monitoring software
CEM uses a sensor validation model as a filter to detect sensor
failures and to set alarms when it identifies a faulty sensor. If a
failed sensor needs repair, the sensor model is also able to act as a
substitute by reconstructing sensor values from the other sensor
values in the plant. This allows the PEMS model to continue to
accurately predict the emissions regardless of the interruption of
replacing a faulty sensor and the associated downtime with the
failed hardware. Fig. 4 shows predictive monitoring software
compares modeled sensor output with actual sensor data for sen-
sor validation.
 PLANT SAFETY AND ENVIRONMENT SPECIALREPORT

Armed with accurate, real-time emissions information, pro-
ducers can identify vulnerabilities before they cause emission limit
violations and continually factor emissions into plant automation
strategies to both reduce emissions and improve efficiency.
Utility, industrial/commercial/institutional (I/C/I) combus-
tors and process heaters (P/H) are fired via a carbon-based fuel
such as natural gas, oil, coal or some form of biomass that is con-
tinuously fed into the combustor chamber. During this process,
complete fuel combustion occurs. However, using these fuels can
be minimized by optimizing burner efficiency, thus reducing fuel
consumption and emissions output.
Efficiency is a direct correlation of data derived from the mea-
surement of flue gas temperature and oxygen (O2). The combus-
tion chamber of an industrial/commercial boiler introduces the
primary air and fuel. The fuel is introduced through a burner
nozzle and is designed to produce a flame front over the full range
of operating conditions. Complete combustion is a function of
oxygen and temperature: the greater the amount of excess oxygen,
the less fuel efficient the boiler will be. Reducing the excess oxygen
improves efficiency. Conversely, as oxygen is reduced, CO can
start to form. Excess levels of CO indicate incomplete combustion
and increased emissions. The use of CO monitoring with a PEMS
can ensure that emission levels remain within limited parameters
and maximize the efficiency of the burner operation. Monitoring
O2 and CO as a process application for burner/boiler control has
been widely accepted in the utility and industrial source market
and has played a strong role in optimizing burner/boiler efficiency
(Fig. 5).
A PEMS is also cheaper. By eliminating the instrumentation
required for a hardware-based approach, this software can predict
emissions based on process variables. Across industries, a PEMS
costs, on average, around 50% less than a hardware-based CEM
system. In industries like oil and gas processing, where hardware
must be able to withstand extremes and function in hazardous
locations, installation and maintenance of a PEMS can cost 70%
less than hardware solutions.
The producer implemented the predictive monitoring software
system and achieved 6% relative accuracy in its first attempt at
RATA certification, well above the EPA regulatory requirement of
+/–20%. Software CEM offers hybrid modeling, through empiri-
cal models and first-principles knowledge, to provide an extremely
realistic representation of process behavior. This approach, com-
pared to other approaches that use a look-up table, helps provide
improved emissions predictions even in the extreme operating
ranges of the unit operations—especially for ultra-low-NOx
burner boilers and turbines.
Unlike historical reporting systems that provide latent informa-
tion from hardware-based CEM systems, software CEM operates
in real time, allowing the plant to monitor operating conditions
that could affect final emissions output. This predictive methodol-
ogy gives the company the ability to simultaneously incorporate
process behavior and feedback into the control strategy of the
boiler and SCR, making it the first of its kind to do so.
In addition to helping prevent fines for noncompliance, the
plant saved more than $100,000 per year in costs relative to main-
taining a hardware-based CEM system. The automated, real-time
and on-demand reporting capabilities of the system saved more in
Alarms
700
600
500
400 Final validated
300 Sensor model signal
200 output
100 Raw sensor
0 measurement
flagged as error
-100
-200
-300
July November

Clean up, comply, cut costs. A large Houston-based mid- FIG. 4 Sensor validation model output and validated signal
stream natural gas company was able to achieve a 90% reduction for failed temperature transducer.
in NOx emissions, using predictive monitoring software along
with ultra-low NOx burner and selective catalytic reactor (SCR)
technologies at one of its natural gas plants. The plant processes
up to 700 MMscf of raw natural gas per day. When its “grand-
fathered” operating limits for NOx emissions expired, the Texas
Commission on Environmental Quality mandated that it reduced Total stack
Carbon monoxide CO, ppm

CO losses at
emissions by 90%. No small task. low low load
CO
The plant installed a new boiler with ultra-low-NO x burn- load
Heat loss, %

high
ers and a SCR. The SCR is an integral component to the boiler load Total stack
exhaust and it uses a specialized catalyst for the reduction of NOx losses at
high load
emissions. The company knew that an accurate and reliable mea-
sure of NOx was critical to the feedback control loop for SCR/
ammonia injection optimization.
Once this equipment was in place, the company had to
choose a CEM system under considerable challenges, including: 300
nonlinear emissions of the combined low NOx boiler with the 50
SCR ammonia injections, inherent low-measurement signal-to- CO control band Excess air
noise levels, and extreme operating conditions within the emission (unaffected by Variation in O2 setpoint
stack. The impact of these challenges meant the potential for boiler load change)
penalties or fines associated with downtime as well as suboptimal
economic control of ammonia usage due to inconsistent and FIG. 5 Representation of the relationship of O2 and CO with
inaccurate feedback measurements. efficiency.

HYDROCARBON PROCESSING NOVEMBER 2010

I 77
SPECIALREPORT PLANT SAFETY AND ENVIRONMENT
reporting costs. Over the project’s life cycle , savings will continue
to grow due to higher reliability and lower maintenance costs of
the software CEM.
Implementing the software. Deploying a PEMS is much
faster than the hardware-based alternatives. Implementing a soft-
ware-based CEM usually begins with discussing project scope.
Participants establish the monitoring system’s functional design
and identify the necessary process sensors. If there is no existing
process in place (i.e., hardware CEM system or other monitoring
system), engineers obtain process data from the process’s data
historian or data logger. If no emissions data is available from an
existing hardware-based CEM systems, an environmental testing
firm can concurrently collect stack emissions data.
The emission unit goes through its entire range of operation
over a two- to seven-day period (depending upon the complexity
of the unit) while data is collected. Engineers use the process and
stack data to construct a highly accurate emissions model.
Beyond compliance. Regulatory monitoring and reporting,
followed by tighter regulatory control, intend to make oil and gas
production “greener.” Processors will need to make investments
in new monitoring and reporting systems to meet regulatory
objectives. Installing a PEMS helps protect a plant from regula-
tory emissions and reporting violations at a fraction of the price
of a hardware-based CEM system. It also provides plants with a
mechanism to better control emissions going forward.
Implementing environmentally friendly approaches to oil and
gas processing and optimizing productivity need not be mutually
Get the Information You Need to Compete in 2011…
Order the Full Version of the
t HPI Market Data 2011
Select 170 at www.HydrocarbonProcessing.com/RS
78
I NOVEMBER 2010 HYDROCARBON PROCESSING
exclusive. By providing plant engineers with accurate predictions
of emissions output in a large variety of circumstances, PEMS
allows these engineers to make strategic emissions management
decisions. This facilitates more efficient processing and gives
plants a leg up in future GHG cap-and-trade markets. PEMS
data and guidance can help oil and gas processors meet goals for
cleaner air, healthier communities and a stronger bottom line.
An emissions data collection is an asset that can make processing
plants more competitive. A PEMS helps make that collection
more accurate and affordable. HP
Richard Hovan has over 34 years of experience in the com-
bustion, safety, environmental instrumentation and air pollution
control equipment field. In addition, he has worked on grass-roots
projects such as an 1,800-MW power plant in Egypt and and several
water-treatment facilities. Mr. Hovan’s background includes such
companies as Graver Water Company, Infilco Degrimont, Inc., Environmental Elements
Corporation. He has been the vice president of product technology for KVB/Analect,
vice president of combustion sales and product marketing for Land Instruments Interna-
tional, markets manager/corporate strategist at Forney Corporation and, most recently,
manager of environmental solutions for Rockwell Automation. He has extensive over-
seas experience with all of his previous companies. Mr. Hovan’s knowledge includes
continuous emissions monitoring (CEMS), boiler and ground-based turbine combus-
tion efficiency monitoring, air pollution control equipment, coal pre-ignition detection
systems and water treatment equipment. He has developed several new processes
and has been instrumental in the development and introduction of new technologies
into the market place. Mr. Hovan has served on the US-EPA’s technical review board
for compliance assurance monitoring (CAM), open market trading rule (OMTR) and
medical waste Incinerators (MWIs) and an opacity performance specification rewrite.
He has developed the first conditional performance specification (CPS) for the EPA. He
is an active member of many organizations.
The HPI’s most comprehensive
comprehens and trusted forecast of capital,
maintenance and operating expenditures for the local and
global HPI, the HPI Market Data 2011 includes over 75 pages
of expanded coverage and detailed analysis for:
• HPI Economics
• Maintenance and Equ
Equipment
• Natural Gas/LNG
• Petrochemicals
• Refining
• En
Health, Safety and Environment
Includes a BONUS CD featuring:
f
• More than 10 years of trends on global construction
activity and spending from previous forecasts.
• HPI Construction Box
Boxscore Updates 2009 and 2010.
• 20 articles from Hydro
Hydrocarbon Processing picked by the
editors to highlight ke
key trends and issues.
ORDER YOUR COPY TODAY and make
strategic decisions in 2011 and beyond!
Online: www.GulfPub.com/2011hpi
Call: +1 (713) 520-4426
 PLANT SAFETY AND ENVIRONMENT
Consider switching to
Internet protocol surveillance
Here’s a checklist to make the jump
M. S. WILSON, Infinova, Monmouth Junction, New Jersey
O
rganizations have experienced a
dramatic demand for surveillance
technology development to protect
people, as well as private and public assets.
The growing need for increasing security,
especially surveillance, leaves many security
managers in a quandary.
A challenging dilemma that petro-
chemical industry security managers face
is how and when they should take the
leap from an analog system to an Internet
protocol (IP)/digital video system. They
want to jump to IP surveillance in a cost-
managed way that extends the existing
equipment life. For most sites, this migra-
tion will take place gradually and, during
the process, analog and IP solutions will
have to coexist, in some cases, for many
years to come.
Traditionally, in the leap from analog
to digital video, organizations convert ana-
log signals to digital signals by buying and
installing rack encoders for their bank of
analog cameras. They replace the analog
control room equipment with new IP con-
trol room equipment. This can be quite
expensive at the front end.
Some believe that a better way is to cre-
ate a coexistent system. In this scheme,
the system keyboards connect to a virtual
memory system (VMS), not the matrix
switchers. The analog side of the coexisting
system stays untouched. Nothing is added
to it. However, since the VMS sits on top of
the system, operators use their traditional
keyboard commands to manage both the
analog and digital solutions.
This is true since the VMS interfaces
with both the system’s analog matrix switch-
ers as well as the IP cameras. As a result, on
the combined video wall, the analog and
IP solutions coexist but are still separate.
Transparent to the operator, with no mouse
needed, the system sends IP camera images
to the digital monitors and analog cam-
era signals to the analog monitors. With
this coexistent solution, agencies can begin
using an IP solution simply by adding IP
cameras, digital monitors and the coexis-
tence VMS.
In the leap from analog to digital, five
major system areas need to be considered:
• Cameras
• Transmission and cabling, including
power supplies
• Storage and retrieval
• Command and control
• Integration.
Cameras—throw out the analog
or keep them. A key consideration for
security professionals is whether or not the
existing analog cameras or new IP ones
will provide the image quality needed to
achieve the functional requirements of the
system. Different applications have differ-
ent requirements; some users require the
ability to see and track suspects in chang-
ing lighting conditions, while others sim-
ply need to see that a corridor is clear.
In many migration plans, specific loca-
tions of greater vulnerability or image detail
requirements are ideal places for IP-based
cameras, including megapixel and high-
definition models, and one needs to ask if
higher-resolution cameras can help at each
location (Fig. 1). A risk/vulnerability matrix
can display overall elements or drill down
to specific locations such as the perimeter,
parking garage, entrance and exit doors,
hallways, computer center, security com-
mand and control (Fig. 2).
Typically, a hybrid approach is consid-
ered in which analog-to-digital encoders
at the camera end can transform images
from an analog camera to digital trans-
mission and storage. The analog control
room equipment gets scrapped but the new
SPECIALREPORT
IP control room equipment controls the
already-installed analog cameras.
Coexistence is a more cost-effective
approach that holds down the budget at
the beginning. The existing analog equip-
ment, including cameras, control room,
video wall and cabling remains untouched.
VMS software, integrated with the present
keyboard, sits on top of the system to man-
age the new IP equipment and the already-
installed analog system.
The petrochemical industry has another
major camera challenge as well—the threat
of explosives. Cameras must be in explo-
sion-proof product housings that meet
stringent corrosion-resistant requirements
(Fig. 3). These housings should be made of
316 stainless steel and be suitable for both
indoor and outdoor installations. The cam-
era should comply with the IECEx standard
and have an IP66 environmental protection
designation. The standard for explosion-
proof camera housing is ExdIICT6.
FIG. 1 High-definition (HD) 36°
continuous rotation megapixel
IP PTZ dome cameras with
1.3-megapixel resolution.
HYDROCARBON PROCESSING NOVEMBER 2010
I 79
SPECIALREPORT
Likelihood
assessment
Threat experts
Reduced
likelihood
Reduced
impact
Vulnerability
Mitigation strategies
Transportation
professionals
FIG. 2 Risk/vulnerability marix.
FIG. 3 Explosion-proof camera housing.
There are also explosion-proof inte-
grated pan tilt zoom (PTZ) cameras that
are designed for use in flammable and
explosive hazardous areas. The housings
on these explosion-proof cameras are
sealed to prevent the intake of explosive
dust while featuring strong corrosion-
resistant capabilities.
Transmission choices—budgets
can dictate. Coaxial, shielded twisted
pair and unshielded twisted-pair cable,
fiber optics and, to a lesser degree, a variety
of wireless approaches carry most security
video. The difference and business advan-
tage of the various transmission schemes
are in the installation and maintenance
costs. A question to ask is whether or not
the new IP cameras will eliminate long dis-
tance analog cabling.
One strategy to handle both analog
and digital networks is to transmit all
the signals over a single fiber optic cable
that is secure and immune to electrical or
environmental interference. Installation is
dramatically simplified by eliminating the
need for multiple fibers, transmitters and
receivers. Not to be forgotten are power
supplies. Following a coexistence plan,
power supplies that are multi-tap, address-
able and programmable have advantages.
Other considerations include the
80
I NOVEMBER 2010 HYDROCARBON PROCESSING
PLANT SAFETY AND ENVIRONMENT
Impact
assessment
Vulnerability
High
Medium
Low
experts
Criticality
Critical
assessment
increased bandwidth impact on the enter-
prise’s network. This is a tricky assignment
and IT can help. Newer types of compres-
sion, decompression or codec, such as
H.264, reduce bandwidth traffic load but
at a cost of more storage and command
center processing. Can the budget afford
the increased transmission and storage
associated with megapixel cameras?
Storage and retrieval challenges.
Though being analog-based, most secu-
rity organizations already have digital and
network video recorders for storage and
retrieval. However, storage solutions have
their own challenges, thanks to myriad fea-
tures and benefits that can range from com-
mon specs to helpful elements such as intel-
ligent PTZ control with preset positions
and e-mail or SMS message notification
upon motion detection or event alerts.
Migrating from MJPEG to H.264 can
reduce storage use by 50% or more. That’s
why security users migrated from MJPEG
to MPEG-4 and now are moving to H.264.
It compresses video into a smaller size, yet
maintains the same video quality when
compared with an MPEG-4. With an
H.264, a representative frame (R-frame)
is selected from a group of frames in a
video sequence. Only the selected R-frame
is stored. By using R-frames, H.264 can
compress a video stream, thereby more
efficiently generating significantly less
bandwidth.
This is true for most camera situations,
such as a fixed camera with a low amount
of motion. If there is a lot of motion, as in
an airport lobby, or if the camera is moving,
such as a PTZ, the number of the R-frames
generated will increase. In some situations,
the compression provided by H.264 may
be only marginally better than M-JPEG.
Nonetheless, because of the lower band-
width generated by H.264, less storage is
required to archive the video. Overall, in
most surveillance situations, H.264 is a
more efficient codec to use for both band-
width reduction and storage.
At the camera edge, security managers
are deploying secure digital or secure digi-
tal storage cards, as well. This is especially
important in applications where connec-
tion loss to the rest of the system could lead
to lost images.
Regardless, there are several questions
to consider before selecting one mode or
another on the pathway to IP:
• If the video is being monitored from
a remote location (and it typically is), will
one get exception reporting?
• Do files ever need to be shared with
other departments, including law enforce-
ment, in real time?
• How much does one need to record
and how long does one need to keep those
recordings?
Command and control options.
There is a lot to consider with command
and control. Traditional matrix switching
and joysticks are workhorses but in a fast-
approaching software world, a solid next
step is to consider networked video matrix
switchers.
Integration. True security systems inte-
gration is a goal of most security operations.
Beyond relays and interfaces, seamless inte-
gration of security video with electronic
access control, intrusion, perimeter, and
identification systems is a beneficial end-
point of any operation and one made sim-
pler through IP.
The bottom line for security
operations. No matter the speed of the
change-over, a solid plan is where both
analog and IP cameras can coexist. Such
coexistence increases security’s overall situ-
ational and domain awareness, improves its
operational effectiveness and efficiencies,
and provides a growth plan that extends the
existing equipment’s life. It also makes sys-
tems affordable and easy to manage. HP
Mark S. Wilson has overall
responsibility for Infinova’s product
marketing and global marketing ini-
tiatives as well as extending relation-
ships with manufacturing partners.
He focuses on building a program structure for global
marketing activities and matrixes with product manage-
ment and manufacturing to develop efficient processes
for new product launches and marketing operations.
 ENGINEERING CASE HISTORIES

Case 59: Heat-up rates

and thermal cracking
A good analysis is usually better than speculation
T. SOFRONAS, Consulting Engineer, Houston, Texas

A
n extruder is analyzed in this case history. While not a This type of analysis obviously tests one of the possible
piece of machinery familiar to most engineers, the tech- causes and does so rather quickly. In any problem-solving ses-
niques used in this analysis can just as well be applied sion, there will be many potential causes from which the most
to pumps, compressors, heat exchangers, vessels or piping. probable will need to be addressed. All proposed causes deserve
The extruder in this example processes polymer. Its main to be tested in some way. Analytical analysis is one way to evalu-
component, a barrel, is essentially a thick pipe with longitudinal ate them. HP
cooling passages in the wall. A screw within the barrel pushes
product through a die at the discharge end. This pushing force
generates a temperature rise due to the shearing action between
the rotating screw and polymer. To keep the temperature constant
at 525°F, water is pulsed through the drilled passages; the water
turns to steam at a moderate temperature, which then helps to
cool the product (the polymer) and keeps it from degrading.

Problem. A malfunctioning control valve allowed water that

was not pulsed to go through the cooling passages at 150°F. There
was concern that the sudden increased temperature difference
could cause cracking between the water passages and the barrel
bore. Figs. 1a and 1b represent a finite element model that has
been reduced to ¼ symmetry and will produce results similar to a FIG. 1 a) Temperature profile at 5 sec.
full model. The longitudinal stress is considered to be negligible, b) Stress profile at 5 sec.
meaning it isn’t restrained axially.
After five seconds, Fig. 1a shows the extruder at its 525°F prod-
uct temperature, except for the cooling passage bore area, which is
at 200°F. Fig. 1b shows that the stress in the bore is 80,000 lbs/in.2
in the x direction (horizontal) and also in the y direction (which is
not shown here). If there were any cracks in the cooling passage,
as illustrated in Fig. 2, this tensile stress would try to open them.
If there was a high tensile stress region going from the bore to the
cooling passage, this would be a concern since a crack could have
propagated through this region.
The tensile strength of the extruder barrel is 140,000 lb/in.2
It is made of ductile material, with no significant tensile stresses
from the passages to the bore. For these reasons, the probability of
a through-crack developing is judged to be relatively low.

Applied results. The importance of this case history is to

show the effect of suddenly applied temperatures and stresses
they produce. Methods, such as presented here, can be used on all
types of heat-up and cool-down situations. The model includes
time, heat transfer coefficients and internal pressure. Of course,
the analysis makes many assumptions. However, if someone
were to ask if the process should now be shut down for a barrel
inspection, you could provide adequate information to say that it
probably has not cracked due to the malfunctioning valve.
FIG. 2 Typical temperature-induced cracking.
Dr. Anthony (Tony) Sofronas, P.E., was worldwide lead
mechanical engineer for ExxonMobil before his retirement. Informa-
tion on his books, seminars and consulting, as well as comments to
this article, are available at http://mechanicalengineeringhelp.com.
HYDROCARBON PROCESSING NOVEMBER 2010
I 81
HPI MARKETPLACE

Select 201 at www.HydrocarbonProcessing.com/RS

WABASH SELLS & RENTS

Call 713-520-4449 for details about
BOILERS & DIESEL GENERATORS
Hydrocarbon Processing’s
FAST EMERGENCY SERVICE
Recruitment
800-704-2002 Advertising
www.wabashpower.com Program
FAX: 847-541-1279 Use a combination of print,
Select 202 at www.HydrocarbonProcessing.com/RS
847-541-5600 recruitment e-newsletter, plus
Website to reach our total
SURPLUS GAS PROCESSING/REFINING EQUIPMENT
NGL/LPG PLANTS: 10 – 600 MMCFD
audience circulation
AMINE PLANTS: 60 – 5,000 GPM of more than 100,000!
SULFUR PLANTS: 10 – 1,200 TPD
FRACTIONATION: 1,000 – 15,000 BPD
HELIUM RECOVERY: 75 & 80 MMCFD
NITROGEN REJECTION: 25 – 80 MMCFD Select 204 at www.HydrocarbonProcessing.com/RS
ALSO OTHER REFINING UNITS
We offer engineered surplus equipment solutions.
Bexar Energy Holdings, Inc.
Phone 210 342-7106
Fax 210 223-0018 Visit our Website at
www.bexarenergy.com
Email: info@bexarenergy.com www.HydrocarbonProcessing.com
Select 203 at www.HydrocarbonProcessing.com/RS

(JWF
ZPVSTFMG
B
DPNQFUJUJWF
BEWBOUBHF
XJUI
SFQSJOUT

$6450.

3&13*/54 $BMM


PS
TBMFT!GPTUFSQSJOUJOHDPN
'PS
BEEJUJPOBM
JOGPSNBUJPO
QMFBTF
DPOUBDU
5BLF
BEWBOUBHF
PG


'PTUFS
1SJOUJOH
4FSWJDF
UIF
PGGJDJBM
SFQSJOU
ZPVS
FEJUPSJBM
FYQPTVSF
QSPWJEFS
GPS
)ZESPDBSCPO
1SPDFTTJOH

82
I NOVEMBER 2010 HYDROCARBON PROCESSING
 Statement of Ownership,
Management and Circulation
(Required by 39 U.S.C.).
1. Title of publication: HYDROCARBON PROCESSING
2. Publication number: ISSN 0018-8190.
3. Date of filing: October 1, 2010
4. Frequency of issue: Monthly.
5. Number of issues published annually: 12
6. Annual subscription price: $199.00.
7. Complete mailing address of known office of pub-
lication: Gulf Publishing Company, P. O. Box 2608
(2 Greenway Plaza, Suite 1020, 77046), Houston,
Harris County, Texas 77252-2608.
8. Complete mailing address of the headquarters or
general business offices of the publishers: Gulf Pub-
lishing Company, P. O. Box 2608 (2 Greenway Plaza,
Suite 1020, 77046), Houston, Texas 77252-2608,
Contact person: Suzanne McGehee; Telephone (713)
529-4301.
9. Names and complete addresses of publisher, edi-
tor and managing editor: Publisher—Bill Wageneck,
2 Greenway Plaza, Suite 1020, Houston, Texas 77046;
Executive Editor—Stephany Romanow, 2 Greenway
Plaza, Suite 1020, Houston, Texas 77046; Director-
Production—Sheryl Stone, 2 Greenway Plaza, Suite
1020, Houston, Texas 77046.
10. Owner: Euromoney Institutional Investor PLC,
Nestor House, Playhouse Yard, London EC4V 5E4,
United Kingdom
11. Known bondholders, mortgagees and other secu-
rity holders owning or holding 1 percent or more of
Bill Wageneck, Publisher
2 Greenway Plaza, Suite 1020
Houston, Texas, 77046 USA
P.O. Box 2608
Houston, Texas 77252-2608 USA
Phone: +1 (713) 529-4301, Fax: +1 (713) 520-4433
E-mail: Bill.Wageneck@GulfPub.com
www.HydrocarbonProcessing.com
SALES OFFICES—NORTH AMERICA
IL, LA, MO, OK, TX
Josh Mayer
5930 Royal Lane, Suite 201, Dallas, TX 75230
Phone: +1 (972) 816-6745, Fax: +1 (972) 767-4442
E-mail: Josh.Mayer@GulfPub.com
AK, AL, AR, AZ, CA, CO, FL, GA, HI, IA, ID, IN, KS,
KY, MI, MN, MS, MT, ND, NE, NM, NV, OR, SD, TN,
TX, UT, WA, WI, WY, WESTERN CANADA
Laura Kane
2 Greenway Plaza, Suite 1020, Houston, Texas, 77046
Phone: +1 (713) 520-4449, Fax: +1 (713) 520-4459
E-mail: Laura.Kane@GulfPub.com
CT, DC, DE, MA, MD, ME, NC, NH, NJ, NY, OH, PA, RI,
SC, VA, VT, WV, EASTERN CANADA
Merrie Lynch
20 Park Plaza, Suite 517, Boston, MA 02116
Phone: +1 (617) 357-8190, Fax: +1 (617) 357-8194
Mobile: +1 (617) 594-4943
E-mail: Merrie.Lynch@GulfPub.com
DATA PRODUCTS AND CLASSIFIED SALES
Lee Nichols
Phone: +1 (713) 525-4626, Fax: +1 (713) 525-4631
E-mail: Lee.Nichols@GulfPub.com
84
I NOVEMBER 2010 HYDROCARBON PROCESSING
total amount of bonds, mortgages or other securities:
Euromoney Institutional Investor PLC, Nestor House,
Playhouse Yard, London EC4V 5E4, United Kingdom
12. Tax Status has not changed
13. Publication title: Hydrocarbon Processing
14. Issue date for circulation data below: September
2010
15. Extent and nature of circulation-Average number
of copies each issue during preceding 12 months. (A)
Total number of copies (net press run)-28,211; (B) Le-
gitimate paid and/or requested distribution; (1) Out-
side County Paid/Requested Mail Subscriptons stated
on PS Form 3541-11,502. (2) In-County Paid/Request-
ed Mail Subscriptions stated on PS Form 3541.)-0-; (3)
Sales through dealers and carriers, street vendors, and
counter sales, and other paid or requested distribution
outside USPS-15,192. (4) Requested Copies distribut-
ed by Other mail classes through the USPS - 46; (C) To-
tal paid and/or requested circulation (sum of 15B1,2,3
and 4)-26,740; (D) Nonrequested distribution (by mail
and outside the Mail). (1) Outside County nonrequest-
ed Copies Stated on PS Form 3541 -0-; (2) In-County
Nonrequested copies Stated on PS Form 3541 -0-; (3)
Nonrequested copies distributed Through the USPS by
Other Classes of Mail -0-; (4) Nonrequested Copies
Distributed Outside the Mail - 815; (E) Total Nonre-
quested Distribution (sum of 15d) - 815; (F) Total dis-
tribution (sum of 15C and E)-27,555; (G) Copies not
distributed -656; (H) Total (sum 15F and G)-28,211;
(I) Percent paid and/or requested circulation (15C/F x
100)-97.04%. Actual number of copies of single issue
SALES OFFICES—EUROPE
FRANCE, GREECE, NORTH AFRICA, MIDDLE EAST,
SPAIN, PORTUGAL, SOUTHERN BELGIUM,
LUXEMBOURG, SWITZERLAND, GERMANY,
AUSTRIA, TURKEY
Catherine Watkins
30 rue Paul Vaillant Couturier
78114 Magny-les-Hameaux, France
Tél.: +33 (0)1 30 47 92 51, Fax: +33 (0)1 30 47 92 40
E-mail: Watkins@GulfPub.com
ITALY, EASTERN EUROPE
Fabio Potestá
Mediapoint & Communications SRL
Corte Lambruschini - Corso Buenos Aires, 8
5° Piano - Interno 7
16129 Genova - Italy
Phone: +39 (010) 570-4948, Fax: +39 (010) 553-0088
E-mail: Fabio.Potesta@GulfPub.com
RUSSIA/FSU
Lilia Fedotova
Anik International & Co. Ltd.
10/2 Build. 1,B. Kharitonyevskii Lane
103062 Moscow, Russia
Phone: +7 (495) 628-10-333
E-mail: Lilia.Fedotova@GulfPub.com
UNITED KINGDOM/SCANDINAVIA,
NORTHERN BELGIUM, THE NETHERLANDS
Michael Brown
1 Ladythorn Crescent
Bramhall
Stockport
Cheshire SK7 2HB, UK
Phone: +44 161 440 0854, Mobile: +44 79866 34646
E-mail: Michael.Brown@GulfPub.com
published nearest to filing date: (A) Total number of
copies (net press run)-26,825; (B) Legitimate Paid and/
or requested distribution; (1) Outside County Paid/Re-
quested mail Subscriptions stated on Form 3541 (in-
clude advertiser’s proof and exchange copies)-10,841.
(2) In-County Paid/Requested mail Subscriptions stat-
ed on PS Form 3541-0; (3) Sales through dealers and
carriers, street vendors, and counter sales, and other
paid or requested distribution outside USPS-14,515.
(4) Requested copies distributed by other mail classes
through the USPS- 39(C) Total paid and/or requested
circulation (sum of 15B1,2,3 and 4)-25,395; (D) Non-
requested distribution (by mail and Outside the mail);
(1) Outside County Nonrequested copies stated on PS
Form 3541 -0-; (2) In-County Nonrequested copies
distributed as stated on PS Form 3541 -0-; (3) Nonre-
quested copies distributed Through the USPS by Other
Classes of mail -0-; (4) Nonrequested Copies Distribut-
ed Outside the Mail - 775; (E) Total Nonrequested dis-
tribution (sum of 15d) -775; (F) Total distribution (sum
of 15C and E)-26,170; (G) Copies not distributed- 655;
(H) Total (sum15F and G)-26,825; (J) Percent paid and/
or requested circulation (15C/F x 100)-97.04%.
16. This statement of ownership will be printed in the
November 2010 issue of this publication. Publication
required.
17. Signature and Title of Editor, Publisher, Business
Manager, Or Owner
I certify that the statements made by me above are
correct and complete.
/s/ John T. Royall, President & Chief Executive Officer
SALES OFFICES—OTHER AREAS
AUSTRALIA—Perth
Brian Arnold
Phone: +61 (8) 9332-9839, Fax: +61 (8) 9313-6442
E-mail: Australia@GulfPub.com
CHINA—Hong Kong
Iris Yuen
Phone: +86 13802701367, (China)
Phone: +852 69185500, (Hong Kong)
E-mail: Iris.Yuen@GulfPub.com
BRAZIL—São Paulo
Alfred Bilyk
Brazmedia Rua Brasilio Machado, 371-601
01230 010 São Paulo SP, Brazil
Phone, Fax: +55 (11) 23 37 42 40
Cell: +55 (11) 85 86 52 59
E-mail: Brazil@GulfPub.com
JAPAN—Tokyo
Yoshinori Ikeda
Pacific Business Inc.
Phone: +81 (3) 3661-6138, Fax: +81 (3) 3661-6139
E-mail: Japan@GulfPub.com
INDONESIA, MALAYSIA, SINGAPORE, THAILAND
Peggy Thay
Publicitas Major Media (S) Pte Ltd
Phone: +65 6836-2272, Fax: +65 6297-7302
E-mail: Singapore@GulfPub.com
PAKISTAN—Karachi
S. E. Ahmed
Intermedia Communications
Karachi-74700, Pakistan
Phone: +92 (21) 663-4795, Fax: +92 (21) 663-4795
REPRINTS
Phone: +1 (866) 879-9144 ext. 194
E-mail: rhondab@FosterPrinting.com
 HPIN WATER MANAGEMENT
LORAINE A. HUCHLER, CONTRIBUTING EDITOR

Huchler@martechsystems.com

Utility water boot camp for process engineers—Part 3

In Part 3, we will discuss the typical responsibilities of a util-
ity process engineer to manage their systems and to interface
with process engineers at all of the production units within the
complex. Plants often assign a newly graduated engineer to the
utility water area. The new engineers lack of plant experience
and technical knowledge of water treatment, coupled with their
lack of authority over the utility water systems in the operating
units create many challenges.
Proactive efforts. Process-side issues demand most of the
daily attention of process engineers, and utility operations are
small and often-overlooked technical areas. Utilities process
engineers should discuss problems with the unit process engi-
neers. Likewise, all process engineers should read the weekly ser-
vice reports from chemical suppliers; discuss nonconformances
with shift supervisors and/or operators; participate in equip-
ment inspections (boilers and heat exchangers); and become
acquainted with the routine testing conducted by the operators
in their unit.
Sometimes, the chemical supplier will offer training for pro-
cess engineers. More often, the service representative will coach
the process engineer on technical issues related to their indi-
vidual process unit. Table 3 summarizes typical requirements
for the new process engineer assigned to utilities.
Summary. Engineers assigned to the water plant can reduce
the risk of failure by understanding the “vital few” issues that can
progress rapidly and have catastrophic consequences, as well as
the problems that develop slowly but have equally severe modes
of failure. Slowly developing problems create a huge risk of
normalization of deviance. Utility process engineers must guard
against allowing this “culture of danger.” New process engineers
should rely on their chemical suppliers for data interpretation
and technical information about controlling corrosion, deposi-
tion and microbiological fouling. Engineers in each process unit
should assume responsibility for the proper operation of the util-
ity water systems in their unit, and collaborate with the chemical
supplier and utility-water process engineer in diagnosing prob-
lems and implementing effective corrective actions. HP
End of series: Part 1, September 2010, and Part 2,
October 2010.
The author is president of MarTech Systems, Inc., an engineering con-
sulting firm that provides technical services to optimize water-related sys-
tems (steam, cooling and wastewater) in refineries and petrochemical
plants. She holds a BS degree in chemical engineering and is a licensed
professional engineer in New Jersey and Maryland. She can be reached at:
huchler@martechsystems.com.

TABLE 3. Typical utility process engineer responsibilities

Action or responsibility Frequency Importance

Interface with service representative • When problems occur • Identify key issues to correct

from chemical supplier • When weekly reports or historical data show • Limit severity of problem

significant nonconformances • Learn about water treatment from chemical supplier representative

• During annual reviews

• Minimum of once per month

Review routine operator testing • Initially when assigned to unit • Partner with chemical supplier to ensure proper

methods, data management and • Annually monitoring procedures

corrective action procedures • Confirm conformance to OEM specifications

Review historical trend data • When problems occur • Limit severity of problem

• When weekly report shows nonconformances

High boiler feedwater total hardness • Poor softener or demineralizer operation • Calcium-rich deposits will cause long-term overheat failures

• Condensate contamination in boiler tubes

• Feedwater heater leak

• Feedwater pump seal leak

86
I NOVEMBER 2010 HYDROCARBON PROCESSING