Journal of Scientific & Industrial Research

440
Vol. 67, June 2008, pp. 440-444 J SCI IND RES VOL 67 JUNE 2008

Numerical study of heat loss from boiler using different ratios of fibre-to-
shell from palm oil wastes
Mohamed Harimi1*, S M Sapuan2, M M H Megat Ahmad2 and Fuad Abas2
1
Centre of Minerals and Materials, School of Engineering and Information Technology, Universiti Malaysia Sabah, 88999
Kota Kinabalu, Sabah, Malaysia
2
Department of Mechanical and Manufacturing Engineering, Universiti Putra Malaysia, 43400 Serdang, Selangor Darul
Ehsan, Malaysia

Received 06 August 2007; revised 04 April 2008; accepted 10 April 2008

This study presents effect of excess air and fibre-to-shell (F/S) ratio on heat losses. Five heat losses are computed
based on ASME standard (heat loss) and STANJAN code (flue gases). Heat loss due to dry flue gas is major heat loss in boiler
and has been found statistically affected by the amount of excess air and F/S ratio, whereas other heat losses are somehow
negligible, except heat loss due to moisture and hydrogen in fuel, which is mainly related to variation of F/S ratio. Boiler
efficiency may reach 85% practically if global optimisation based on excess air and F/S ratios are consideredM with respect to
combustion efficiency (low CO) and thermal efficiency (less heat losses).

Keywords: Dry flue gas, Excess air, Heat losses, Palm oil wastes, Ratio of fibre-to-shell

Introduction
Table 1—Proportion of fibre and shell used in the
Most of the palm oil mills in Malaysia use fibre and palm oil mills of Perak1
shell (F/S) as fuel for boiler, but ratio or proper mixing
of these two wastes are not considered. Husain et al1 Mill # Proportional weight, % Calorific value
found that boiler efficiency (73%) affects effective Fibre Shell MJ.kg-1
utilization of steam in palm oil mill, due to non-
1 64 36 14.310
homogeneity of fuel, improper blending of F/S, high 2 60 40 14.621
percentage of moisture and incomplete combustion. The 3 67 33 14.077
proportional weights of F/S used to feed a boiler in seven 4 70 30 13.844
palm oil mills in Perak state of Malaysia (Table 1) show 5 60 40 14.621
6 50 50 15.399
really inconsistency of F/S ratio. 7 50 50 15.399
Palm oil mills are continuing using only F/S (65:35) HHV (Fibre)=11.512 MJ.kg-1; HHV (Shell)=19.285
as fuel for grid-connected power generation and CHP2. MJ.kg-1; Moisture = shell 10%, fibre 40%
Mahlia et al3 highlighted that composition of fuel in
boiler is changing and therefore excess air (EA) used to and 70:30 with an optimal air fuel ratio of 12/1 to 16/1
combust F/S has to be identified, in order to get an respectively. Air fuel ratio of 12/1 corresponds to
acceptable emission limit. Generally, if fuel composition percentage EA > 120%, whereas 16/1 corresponds to
of fuel is not varying too much then the best excess air percentage EA > 150%. Such range may be acceptable
can be obtained and maintained. But if the fuel for reducing black smoke but it will also reduce boiler
composition is changing, then excess air should be efficiency.
changed. Mahlia et al3 used F/S (70:30) with 30% excess This study presents effects of amount of EA and F/S
air without indicating the reason of choosing such ratio on heat losses due to dry flue gas, moisture in fuel,
percentage of EA. Leong4 has suggested F/S of 60:40 incomplete combustion, water from combustion of
*Author for correspondence hydrogen (H2) in fuel, and finally moisture in combustion
E-mail: harimi2000@hotmail.com air.
 HARIMI et al: EFFECT OF EXCESS AIR AND FIBRE TO SHELL RATIO ON HEAT LOSSES
Methodology
Chemical composition of fibre and shell (% by
mass), taken from a palm oil mill in Malaysia, is as
follows: C, 30.68, 47.16; H, 3.90, 5.67; S, 0.20, 0.18;
N, 0.91, 0.54; O, 23.86, 33.57; H2O (moisture), 35.00,
10.00; and ash, 5.45, 2.88%; and higher heating value,
11.69, 18.05 MJ/kg. Fixed data for heat losses,
including average ambient air temperature (Ta), average
relative humidity of air (RH) and stack flue gas are
28°C5-6, 80%5-6, and flue gases (STANJAN Code)17
respectively. Based on the data of RH (80%), amount
of water in air at 28°C is computed using Eq. (1) or
using Psychrometric chart based on a barometric
pressure of 1.01325 bar7.
ω=0.622 Ps / (P – Ps) …(1)
where ω, amount of water vapour per amount of dry air
(kg water/kg air); P, total pressure of water vapour and
air (1.01325 bar); Ps, partial pressure of water vapour
in mixture (bar).
Total pressure P includes Ps and partial pressure of
dry air (Pa), as
P = Ps + Pa …(2)
Amount of fraction of RH is related by
φ = Ps /Pg …(3)
where φ, fraction of RH; Pg, saturation pressure of water
vapour at Ta (bar).
Combining Eqs (1) and (3)
ω = (0.622 φ . Pg) / (P – φ .Ps) …(4)
From steam table at Ta = 28°C => Pg = 0.03778 bar,
and with φ = RH /100 amounting to 0.8 and including
the total pressure of 1.01325 bar, when all those data
are substituted into Eq. (4) gives ω = 0.01912 kg
vapour/kg dry air.
Computation of heat losses is based on ASME
standard and limited to heat losses due to dry flue gas,
moisture in the fuel, incomplete combustion, water from
combustion of hydrogen in fuel, and finally moisture
in combustion air.
441
Heat Losses based on ASME Standard
There are two methods (direct method and indirect
method) of analysing boiler efficiency. In this study,
indirect method is used based on ASME standard PTC
4.18.
Heat Loss due to Dry Flue Gas
There is an energy loss associated with N2, which enters
boiler as a constituent of combustion air and leaves boiler
at a higher temperature. Additionally, gaseous combustion
products leave boiler at an elevated temperature, thus
energy is lost from the system. This loss in a correctly
operated system can be calculated as
HLg = mg . Cpg . (Tg – Ta) …(5)
where HLg, heat loss due to dry flue gas (MJ/kg); mg, mass
of dry flue gas per kg of fuel as-fired; Cpg, average specific
heat capacity of flue gas (0.001 MJ/kg °C between
20-200 °C); Ta and Tg, average inlet air and exit flue gas
temperatures respectively (°C).
Heat Loss due to Moisture in Fuel
Moisture entering boiler with fuel leaves as
superheated vapour. This moisture loss is made up of
sensible heat to bring moisture to boiling point [Cpw .(100
– Ta)], e superheat required to bring this steam to
temperature of exhaust gas [Cpv .(Tg – 100)], and latent
heat of evaporation (hfg). All the three forms of losses are
grouped as
mc f
HLmf =
100
[
⋅ Cp w ⋅ (100 − Ta ) + Cp v ⋅ (Tg − 100 ) + h fg ]
…(6)
where HLmf, heat loss due to moisture in fuel, MJ/kg; mcf,
moisture content of fuel, % as-fired; hfg, latent heat of
evaporation of water (2.26 MJ/kg at 100°C); Cpw, average
specific heat capacity of water (0.0042 MJ/kg°C at
100°C); Cpv, average specific heat capacity of water
vapour (MJ/kg°C).
Heat Loss due to Incomplete Combustion
Product formed by incomplete combustion could be
mixed with oxygen and burned again with further release
of energy. Such products include carbon monoxide (CO),
H2 and various hydrocarbons and are generally only found
in flue gases. CO is the only gas whose concentration
can be determined conveniently in a power plant test9.
The expression of heat loss due to incomplete combustion
has been proposed by Ganapathy10 and Kumar & Sah11
as
442 J SCI IND RES VOL 67 JUNE 2008

Table 2—Percentagea of heat losses due to various parameters

F/S Parameter Excess air, %

ratio 0 20 40 60 80 100 120 150

0:100 HLg 4.38 5.18 6.00 6.81 7.63 8.45 9.26 10.49
HIC 4.6718 1.6041 0.4408 0.1119 0.0288 0.0077 0.0022 0.0003
HLa 0.1566 0.1879 0.2193 0.2506 0.2819 0.3132 0.3445 0.3915
HCL 15.4768 13.2476 12.9278 13.4457 14.2113 15.0397 15.8842 17.1577
20:80 HLg 4.40 5.21 6.03 6.85 7.67 8.49 9.32 10.55
HIC 4.3031 1.3972 0.3705 0.0928 0.0238 0.0064 0.0018 0.0003
HLa 0.1575 0.1890 0.2205 0.2520 0.2835 0.3150 0.3464 0.3937
HCL 16.0639 14.0009 13.8225 14.3966 15.1810 16.0178 16.8679 18.1489
40:60 HLg 4.43 5.24 6.06 6.89 7.72 8.54 9.37 10.61
HIC 3.9014 1.1833 0.3017 0.0746 0.0191 0.0051 0.0014 0.0002
HLa 0.1584 0.1900 0.2217 0.2534 0.2850 0.3167 0.3484 0.3959
HCL 16.7711 14.9003 14.8720 15.5016 16.3044 17.1496 18.0055 19.2939
50:50 HLg 4.44 5.26 6.08 6.91 7.74 8.57 9.40 10.64
HIC 3.6874 1.0744 0.2682 0.0659 0.0168 0.0045 0.0013 0.0002
HLa 0.1588 0.1906 0.2223 0.2541 0.2859 0.3176 0.3494 0.3970
HCL 17.1810 15.4176 15.4675 16.1243 16.9362 17.7854 18.6441 19.9362
60:40 HLg 4.45 5.27 6.10 6.93 7.76 8.59 9.43 10.68
HIC 3.4648 0.9652 0.2358 0.0575 0.0147 0.0039 0.0011 0.0002
HLa 0.1593 0.1911 0.2230 0.2548 0.2867 0.3185 0.3504 0.3981
HCL 17.6371 15.9895 16.1188 16.8023 17.6227 18.4761 19.3377 20.6337
80:20 HLg 4.48 5.31 6.14 6.97 7.81 8.65 9.48 10.74
HIC 3.1682 0.6811 0.1591 0.0383 0.0098 0.0026 0.0007 0.0001
HLa 0.1602 0.1922 0.2242 0.2563 0.2883 0.3203 0.3524 0.4004
HCL 18.9592 17.0773 17.8030 18.6330 19.4884 20.3508 21.1651 22.4689
100:0 HLg 4.51 5.34 6.18 7.02 7.86 8.70 9.54 10.81
HIC 2.4805 0.5404 0.1205 0.0287 0.0073 0.0020 0.0006 0.0001
HLa 0.1611 0.1933 0.2256 0.2578 0.2900 0.3222 0.3544 0.4028
HCL 20.0924 19.0147 19.4648 20.2455 21.0980 21.9670 22.8401 24.1518
a
Percentage of heat losses are computed with respect to higher heating value of F/S (fibre to shell ratio)
HLg, computed based on Eq. (5); HIC, computed based on Eq. (7); HLa, computed based on Eq. (9); HCL, computed
based on Eq. (10)

9 ⋅ mH 2
HIC = HHVCO ⋅ mC ⋅
y2
…(7)
HLH =
100
[
⋅ Cp w ⋅ (100 − Ta ) + Cp v ⋅ (T g − 100 ) + h fg ]
y1 + y 2
…(8)
where, HLH, heat loss due to water from combustion of
where HIC, heat loss due to incomplete combustion (MJ/ hydrogen in fuel (MJ/kg); mH2, hydrogen content in fuel,
kg); HHVCO , higher heating value of CO, typically % as-fired.
HHVCO =10.143 MJ/kg; mC, carbon in fuel, % as-fired;
y1 and y2 are volume fraction of CO2 and CO respectively Heat Loss due to Moisture in Combustion Air
in flue gas. Vapor as humidity in incoming air is superheated as
it passes through boiler. Water vapor mass that air
Heat Loss due to Water from Combustion of Hydrogen in Fuel contains can be obtained from Psychrometric charts, and
Combustion of hydrogen causes a heat loss because for this loss Eq. (9) is used.
water is formed in combustion product. This water is
converted to steam in boiler and carries away heat, HLa = ω . ma . Cpv . (Tg – Ta) …(9)
particularly because of its latent heat content. Heat loss where HLa, heat loss due to moisture in combustion air
is calculated as (MJ/kg); ω, mass of water vapor per kg of combustion
 HARIMI et al: EFFECT OF EXCESS AIR AND FIBRE TO SHELL RATIO ON HEAT LOSSES 443

air; ma, mass of combustion air per kg of fuel as-fired; Thus, validity of results obtained using ASME, PTC4.1
Cpv, average specific heat capacity of water vapor (0.002 standard heat losses are valid with the validity of results
MJ/kg°C between 20-200°C). obtained from STANJAN software17 flue gases, and also
Due to some difficulties in estimating some losses related to properties of air, which were taken as an
such as heat loss due to carbon-in-ash, heat loss due to average value5,6 of temperature (28°C) and humidity
sensible heat in fly ash, bottom ash are not included (80%).
though, most of biomass has a low ash content, and Heat losses due to combustion, flue gas, incomplete
represents a small energy loss if dumped hot12-13. Heat combustion and moisture in air have been computed
loss due to radiation and convection is not included, and (Table 2). Variation of heat losses due to moisture and
taken as a maximum value of 2% from energy content hydrogen in the fuel related to given change of F/S ratio
of solid fuel14-15. Sum of five losses can be grouped as but not to the variation of excess air (Table 3). Now in
total heat loss due to combustion (HCL) as order to study the effect of excess air and F/S ratio for
each single heat loss, a statistical analysis is carried out,
HCL = HLa + HLg + HLmf + HLH + HIC …(10) where ANOVA 2-way is used to assess heat losses due
to total heat loss, dry flue gases, and incomplete
Results and Discussion combustion (Table 4). For cases of other losses,
Application of ASME method, PTC 4.1 is mainly statistical t-test is used because they are not related to
based on analysis of flue gases and to any type of fuel the variation of excess air.
used and does not require measurement of the boiler16. Total heat loss, heat loss due to dry flue gases, and
heat loss due to moisture in air are found statistically
Table 3—Percentageb of heat loss due to moisture and highly affected by excess air, and fuel mixture (F/S ratio).
hydrogen in the fuel As for the case of heat loss due to incomplete
combustion, it is highly affected by variation of excess
F/S 0:100 20:80 40:60 50:50 60:40 80:20 100:0 air but not too much affected by F/S ratio. For heat loss
HLmf, % 1.48 2.38 3.43 4.03 4.68 6.16 7.97 due to moisture and hydrogen in the fuel, t-tests
HLH, % 2.80 2.82 2.85 2.87 2.88 2.91 2.97 computed are found to be 5.1323 and 4.9054
b
respectively. From statistical table, critical t-test (degree
Percentage of heat losses are computed with respect to higher
heating value of F/S (fibre to shell ratio)
of freedom, df = 6) is lower than t-tests computed;
HLmf, computed based on Eq. (6); HLH, computed based on Eq. (8) therefore, heat loss due to moisture and hydrogen is more

Table 4—Statistical analysisc of heat losses

Heat Analysis of Excess air, % F/S ratio Error
losses variance
HLg SS 219.2827 0.2718 0.01972
MS 31.3261 0.04530 0.000469
Fcomp 66726.1157 96.4798 -
HIC SS 79.9579 1.1713 2.9250
MS 11.4226 0.1952 0.06964
Fcomp 164.0176 2.8031 -
HLa SS 0.3229 0.000332 0.000026
MS 0.04612 0.000055 6.22 10-7
Fcomp 74177.98 88.93 -
HCL SS 121.1882 239.7505 3.4661
MS 17.3126 39.9584 0.08253
Fcomp 209.7857 484.1970 -
c
Degree of freedom: excess air, 7; F/S ratio, 6; error, 42; total, 55 HLg, heat loss due dry flue gas based on Eq.
(5) and Table 2; HIC, heat loss due incomplete combustion based on Eq. (7) and Table 2; HLa, heat loss due
moisture in air based on Eq. (9) and Table 2; HCL, heat loss due combustion based on Eq. (10) and Table 2
444 J SCI IND RES VOL 67 JUNE 2008
affected by variations of F/S ratio. Boiler efficiency
(100, total heat losses) varies from 76% to a maximum
value of 87% (Table 2). Practically, boiler efficiency
for modern waste fuel-fired boilers used for biomass-
based power plant varies from 80 to 85%20-22.
Conclusions
Statistical method, ANOVA 2-way, was used and
found that heat loss due to dry flue gases and moisture
in air are statistically highly affected by change in excess
air, and fuel mixture (fibre and shell). Heat loss due to
incomplete combustion was found highly affected by
variation in excess air, but not too much affected by F/S
ratio, and it is negligible above 40% excess air. The
t-test confirmed that heat loss due to moisture and
hydrogen in fuel was found more affected by variations
of F/S ratio. Maximum heat loss due to moisture in air
was found to be 0.4% from total heat loss. Boiler
efficiency was found to vary from 76% to a maximum
of 87%, and improvement can be achieved if global
optimisation based on excess air and F/S ratio are
considered with respect to combustion efficiency (low
CO), and thermal efficiency (less heat losses).
References
1 Husain Z, Zainal Z & Abdullah Z, Briquetting of palm fibre
and shell from processing of palm nuts to palm oil, Biomass &
Bioenergy, 22 (2002) 505-509.
2 Pinto F, Malaysia Biomass-based Power Generation and
Cogeneration Palm Oil Industry Project (United Nation
Development Program, New York) 2002.
3 Mahlia T M I, Abdulmuin M Z, Alamsyah T M I & Mukhlishien
D, An alternative energy source from palm wastes industry for
Malaysia and Indonesia, Energy Convers & Manage, 42 (2001)
2109-2118.
4 Leong W L, Control of black smoke and particulates emission
for palm oil mill boilers, Proc Nat Workshop on Recent
Developments in Palm Oil Milling Technology Pollution
Control (PORIM, Kuala Lumpur) 1986.
5 Abdulah A & Aliash H, Generalized long duration probable
maximum precipitation (PMP) isohyetal map for peninsular
Malaysia, —J Spacial Hydrology, 4 (2004) 1-15.
6 Keizrul A, Hydrology for sustainable water resources planning,
development and management in Malaysia, Proc Int Conf
Hydrology and Water Resources of Humid Tropics (Ipoh,
Malaysia) 1998, 1-22.
7 Eastop T D & McConkey A, Applied Thermodynamics for
Engineering Technologists, 5 th edn (Pearson Education
Limited, Singapore) 1993.
8 ASME, Steam generating units, ASME/ANSI Standard PTC
4.1 1974, Reaffirmed 1991, (American Society of Mechanical
Engineers, New York) 1991.
9 ERI, Guide Book 2: Boiler and furnaces. Prepared for the
European Commission DG TREN (Energy Research Institute,
University of Cape Town, South Africa) 2000, 71-85.
10 Ganapathy V, Steam Plant Calculation Manual, 2 nd edn
(Marcel Dekker, Inc., New York) 1994.
11 Kumar A & Sah G N, Thermal Engineering (Narosa Publishing
House, Alpha Science International Ltd, UK) 2004.
12 NRCAN, Les Petites Installations de Chauffage à la Biomasse:
Guide de l’Acheteur, (Natural Resources Canada, Energy
Resources Branch, Canada) 2000, 9-10.
13 Kong H W, Current Status of Biomass Utilization in Malaysia
(Forest Research Institute Malaysia, Kepong, Malaysia) 2000.
14 Boon K Y & Azman F S, Boiler design for palm waste firing,
Proc 1987, Int Oil Palm/ palm Oil Conf (PORIM, Bangi,
Malaysia) 1989.
15 Kock J D, Eliminating dark smoke from palm oil mill boiler
with fuel conditioning and control system, Proc 2001, PIPOC
Int Palm Oil Congr (MPOB, Bangi, Malaysia) 2001.
16 Anke H, Lambert S, & Natalia V, Improving Energy in Peruvian
Boilers with CDM: Feasibility Study for a Bundled CDM
Project (Institute for Applied Ecology, Denmark) 2003.
17 Reynolds W C, STANJAN Program: The element potential
method chemical equilibrium analysis (Thermo-sciences
Division, Department of Mechanical Engineering Stanford
University, California) 1986.
18 Harimi M, Application of theoretical combustion analysis in
determining the optimum fibre/shell ratio for oil mill boiler,
Ph D Thesis, Universiti Putra Malaysia, Serdang, Selangor,
Malaysia, 2006.
19 Harimi M, Hamdan M M, Sapuan S M & Azni Idris, Numerical
optimisation of excess air with respect to fibre-to-shell ratio
during incineration process, Int Energy J, 7 (2006) 125-136.
20 Nicholas L, Power generation from palm oil mill: Design
options and costing. Proc 2002 Nat Semin on Palm Oil Milling,
Refining Technology and Quality and Environment (MPOB,
Kota Kinabalu, Sabah, Malaysia) 2002.
21 Chee T H, New generation boiler for the palm oil industry,
Proc 2000 Nat Semin on Palm Oil Milling, Refining
Technology, Quality and Environment (MPOB, Genting
Highlands, Pahang, Malaysia) 2002, 17-21.
22 Chee T H, Modern boiler systems for the palm oil industry,
Proc 2001 PIPOC Int Palm Oil Congr: Cutting Edge
Technology for Sustained Competitiveness (MPOB, Kuala
Lumpur, Malaysia) 2001, 221-222.