Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Abstract
Production logging plays an important role in understanding the impact that well trajectory, completion practices
and trials, and well placement have on EUR. The data can further help to characterize the reservoir in new plays.
This is particularly true in the case when both water and gas production deviate dramatically from expectations.
Logging helps to answer the questions: Are all zones and perforating clusters producing equally? Is the toe
unloading? What is the flow regime? Is water production isolated to a particular interval? Can it be shut off?
In traditional vertical wells, electric line logging is most often used to acquire production data. In a horizontal well,
both active and memory tools can be deployed on coil tubing to acquire production data. Using coil tubing for
production logging has its own unique challenges, one of which is getting to TD without damaging the tools. The
difficulty of this challenge can be compounded by long lateral lengths and high dog leg severity.
Well trajectory, well flow regime, water rate, and gas rate all have an effect on choice of logging tools and method
for deployment. This paper will discuss two production logging attempts: 3 successful logs with active tools and 1
failure with memory tools. The one failed attempt involved a low gas rate, low pressure, high water rate well with
severe dog legs near the heel where the coil could not reach the logging interval.
The sections in this paper will discuss the role of an effective cleanout and how to achieve its success as well as
best practices for acquiring high quality production data. Each of the four logging jobs will be discussed in detail
with results provided.
SPE 154221
Logging the well under normal flowing conditions provides the most representative data. If excessive fluid is
introduced during the cleanout, the well may require more time to return to normal flowing conditions.
Pumping high quality nitrogen foam during the cleanout can help to maintain well pressure and keep the well
flowing. Targeting 80 quality foam will still lift the solids and debris out of the well and ensure that even a lower
pressure Marcellus well will keep flowing. This was very successful in three of the wells later described. Previous
cleanouts on a typical 9000 MD Marcellus well with a 4000 lateral have required between 500,000 and 600,000
scf of nitrogen. Depending on the service provider, two transports may be required and therefore should be
available on-site. In the case of Well A, sufficient nitrogen was not available on site, and therefore the quality of
the cleanout was compromised. Due to the large amount of nitrogen being pumped, the returns will foam
significantly. Having de-foamer on location with a method of application is essential to control returns and prevent
spills.
A wash nozzle was used for the cleanout runs described in this paper. A drawing of the BHA can be seen below:
Other tools and BHAs will work, but this BHA was very successful in lifting fill and debris out of the wells. Unlike a
wash nozzle, other tools require a motor, which may not be compatible with the high quality nitrogen foam. The
wash nozzle is also relatively low cost compared to other options. If the initial coil tubing plug drillout was
thorough, the wash nozzle should be sufficient. Otherwise, a near drift motor and mill should be used.
For the cleanouts described in this paper, the flowback rig up included the following:
SPE 154221
3iron was used coming off of the tree to prevent any plugging during the cleanout. In all cases, the well had been
previously on production through a sales line. Therefore, a flare was rigged up to handle the gas and nitrogen in
the returns, which cannot be sent into the sales line.
Once all necessary equipment was rigged up and the coil tubing and BOPs were pressure tested with water, the
fluid was pumped out of the reel to an open top tank and replaced by 80 quality N2 foam. This may seem trivial
but has proved very successful. This helped to prevent prematurely killing the well at the beginning of the clean
out run.
The successful cleanout runs have been executed based on the following guidelines:
- RIH @ +/-90fpm.
- Minimize pump rates while RIH unless obstruction is tagged or CT weights are suspicious. Only
introduce fluid into the well when necessary for cleaning up the wellbore.
- Perform weight checks every 2500.
- Upon reaching a depth of 500 above the kick off point, begin pumping 70-80 quality foam.
For the wells described in this paper, downhole conditions dictated the following pump rates for 80Q N2 foam to
establish circulation:
Fluid Rate
0.25 bbl/min
N2 Rate
600 scf/min
During the cleanout, bbl in = bbl out was maintained to aide in circulation. It is crucial to keep the well choked back
as much as is reasonably practical. If the well is allowed to flow excessively, it can bring in more sand and debris,
potentially sticking the coil tubing in the lateral.
At the kick off point, the RIH speed was reduced and pump rates were increased to begin washing. See target
fluid rates (based on estimated reservoir pressure) for washing below:
Target Fluid Rate
Target N2 Rate
CT RIH Speed
0.5 bbl/min
1650 scf/min
35 fpm
The lateral was washed in two sections with a short trip back to the vertical in between. After washing to TD, a
bottoms up circulation was performed. The coil tubing was slowly pulled out of the well. It is important to not
outrun the fluid. See example annular velocity calculations for 2 CT inside 5-1/2 20# casing:
Equivalent Pump Rate (bbl/min)
1.5
2.0
2.5
3.0
SPE 154221
The exact steps of a cleanout will vary from well to well. However, these guidelines have been implemented on all
successful cleanouts and will continue to be used.
After the cleanout is finished, the well should be flowed back. The goal of this flowback is to remove nitrogen from
the wellbore and return the well to normal flowing conditions. The duration of this flowback is a bit unknown and
will be well specific. For the wells described in this paper, the flowback time ranged from 1-3days. This duration,
however, was driven more by operational constraints and facilities coordination than actual well flowing conditions.
Despite the sub-optimal flowback practices executed, the log quality did not suffer.
Logging Procedure
A coil tubing unit should then return to location to perform the logging passes. The logging company and type of
tool selected will dictate the following steps. For the wells described in this paper, both active and memory logging
tools on coil tubing were used. The following practices were followed on all successful logging operations.
After the coil tubing unit was rigged up, the coil tubing was pressure tested with nitrogen for production logging.
This reduced the volume of fluid introduced into the well. During the logging, it is important to keep the well flowing
and limit any excessive disturbances to the well (i.e. pumping unnecessary fluid).
Once the coil tubing was pressure tested, the logs were run using the following guidelines:
- RIH Speed @ +/- 120fpm in the vertical
- Reduce the RIH speed to 90fpm through the deviated section
- Reduce RIH speed further to initial logging speed (usually 30 fpm) 200 above the logging interval.
- The logging interval is described as 200 above the top perforation to PBTD. Begin logging and
perform multiple passes as described in the procedure.
Each well may be different, but the following is how the logs in this paper were executed:
o Log down and then up@ 60fpm
o Log down and then up @ 90fpm
o Log down and then up @ 120fpm
o Log down @ 30fpm if time permits. This pass is important for temperature data.
Two minute stationary stops were performed between all stages during the first logging pass down. During the job,
flowing well head pressure, flowing gas rate, and flowing water rate were recorded every 15 minutes.
Results
All wells discussed in this section were hydraulically fractured with slickwater and conventional plug and perf
methodology. Proppant selection, proppant concentration, and stage spacing were varied from well to well, but
remained consistent within a wellbore. The execution of such designs, however, varied slightly from stage to stage
due to operational challenges. Any dramatic deviations from plan to execution than may have affected production
will be discussed during the sections below. Production logs were run as previously described in this paper. The
results are below:
Well A
Casing: 5-1/2 20# P110 casing
KOP: 4660 ft
Landing: 5900 ft
TMD: 9607 ft
TVD: 5415 ft
Goal: Run a production log with memory tools and understand the gas distribution
Result: Data was not captured for this well. During the cleanout, high quality foam was not maintained due to lack
of nitrogen availability. The coil tubing was pressure tested with water and then the was not pumped out of the coil
prior to RIH. This introduced even more fluid into the well. The well was killed initially and returns were lost during
the cleanout. Excessive fill was encountered in the heel of the well. Memory logging tools were run 2 days after
the cleanout and could not pass through the heel. The producing interval could not be logged. This attempt was
not successful due to a sub-optimal cleanout.
SPE 154221
Well B
Casing: 5-1/2 20# P110 casing
KOP: 3026 ft
Landing: 4983 ft
TMD: 8248 ft
TVD: 4688 ft
Water Contribution
1
2
3
4
5
6
7
8
9
10
11
12
13
14
4.90
11.90
14.70
10.50
7.70
3.50
2.80
13.30
7.00
9.80
3.50
6.30
4.20
trace
Trace
Trace
Trace
Trace
-
% of Perfs
Contributing
25
50
100
100
50
25
50
100
75
75
50
25
75
Could not determine
SPE 154221
Landing: 5834 ft
TMD: 9940 ft
TVD: 5489 ft
Deepest perforation: 9780ft
SPE 154221
5.92
15
9.56
0
0.00
0
6.37
0
8.17
40
7.60
3
7.97
12
3.27
0
7.28
0
9.17
0
1.33
0
6.85
0
4.74
0
3.76
0
2.75
0
1.78
0
0.33
0
4.42
0
8.73
30
Figure 7: Well C Production Log Distribution Reults
% Perf Clusters
Contributing
75
66
0
66
100
25
75
25
66
40
66
100
66
100
66
33
33
66
33
SPE 154221
Based on the results of this log, all stages except stage 3 were contributing to the gas production at the time of the
logging run. During the frac, a sharp drop in pressure indicated that the plug may have slipped during this stage.
Stage 3 was pumped to completion. The proppant intended for stage 3 could have been pumped into stage 2
below, resulting in a lack of production from stage 3.
During the breakdown of both stages 6 and 10, design rate could not be reached at a suitable treating pressure.
As a result two perforation clusters were added above these stages to aid with injection. After these perforations
were added, the frac jobs were pumped to completion. In both cases, the only perforations that contributed to the
gas production were the additional clusters added after the first breakdown attempt. There are a few other stages
where only the shallowest perf clusters contributes. These were not re-perforated stages.
Well D
Casing: 5-1/2 20# P110 casing
KOP: 4489 ft
Landing: 5834 ft
TMD: 9458 ft
TVD: 6780 ft
Deepest perforation: 9300ft
SPE 154221
0.30
0
0.85
0
0.00
0
0.26
0
0.00
0
1.26
0
6.35
0
6.48
0
12.16
6
7.42
14
11.87
16
22.35
23
22.89
11
7.82
30
Figure 10: Well D Production Log Distribution Results
% Perf Clusters
Contributing
25
75
0
25
0
50
100
50
75
50
50
75
75
75
10
SPE 154221
Summary
The first campaign of production logs in the Marcellus yielded positive and promising results.
Both failed and successful logging attempts continue to highlight the important role of the cleanout run. These
practices will continue to be carried on to other production logging projects. Future logging attempts will include a
plan to improve the flowback process between the cleanout and the logging run. Field coordination and facility
constraints have dictated this process on past attempts. In the future, these stumbling blocks will be removed,
further improving the quality of data.
The logging results discussed in this paper provided key insight into the effect that completions trials have on
production. For example the varying perforation strategy between Wells C and D did not seem to affect the % of
perforation clusters contributed. However, within well C, when perforation clusters were added post breakdown
and the frac was pumped to completion, the additional perforations seemed to dominate the production
contribution for that stage. Both of these require further investigation and experimentation.
SPE 154221
11
The logging results showed that the wellbore trajectory can significantly affect the ability of the well to lift liquids to
surface, especially in Wells B and D which were toe down. These results begin to paint that picture that one size
does not necessarily fit all when it comes to artificial lift. There may be value in a more well-specific approach.
In every successful log run, the shallowest stage contributed to the gas production. Because of these results, the
completion plan will continue to including the section of the well directly above the heel, despite the deviation being
less than 75 degrees.
Moving forward, production logs will be run in conjunction with other technology trials to better understand how frac
placement, well location, and wellbore trajectory affect production.
Acknowledgements
This project would not have been possible without the hard work and dedication of the following people:
Craig Wasson and Mark Miller, SLB Coil Tubing Horseheads, NY
Chris McCann and Quincy Smith, SLB Wireline Graham, TX
Alex Menkhaus, SLB CoilTOOLS (Active Tools)
Beau Woodward , Shell On-Site Supervisor
John Harrod, Shell Production Technologist