- Email: [email protected]
Experimental investigation aiming at improving the suction flow capability of a gas expeller
Journal of Natural Gas Science and Engineering 23 (2015) 458e463
Contents lists available at ScienceDirect
Journal of Natural Gas Science and Engineering journal homepage: www.elsevier.com/locate/jngse
Experimental investigation aiming at improving the suction flow capability of a gas expeller Kamal K. Botros a, *, John Geerligs a, Brad Watson b a b
NOVA Chemicals Centre for Applied Research, Calgary, Alberta, Canada TransCanada Pipelines Limited, Calgary, Alberta, Canada
a r t i c l e i n f o
a b s t r a c t
Article history: Received 14 January 2015 Received in revised form 18 February 2015 Accepted 19 February 2015 Available online 3 March 2015
An expeller performance has been evaluated in terms of its capability to induce higher suction flow for application to evacuate combustible gases from a blown down natural gas pipelines. The investigation involved a test rig and testing of a typical 150 mm (nominal size) expeller. This particular expeller has 12, 2.35 mm diameter holes, equally spaced around the throat circumference of the expeller. This was referenced to as the base (or original design). The aim of the present investigation is to improve the suction flow capability of this expeller by four modifications to both the number and/or sizes of these holes. The experimental results showed that the performance of the expeller in terms of its capability of driving higher suction flows for a given flow resistance system can be improved by increasing the number and sizes of the drive air holes which in turn permit higher drive air flow. However, with increased drive air flow, the performance of the expeller in terms of the induction ratio (IR) deteriorates, but luckily not at the same rate as the suction flow increases. Hence a cost effective means to improve the suction flow capability of an expeller is to drill more and larger size holes around its throat. The loss in the IR (which is efficiency related), however, is generally not a concern in practice when the economic benefit of evacuating the pipeline section in a timely and safe manner greatly overweigh any potential loss in the expeller IR efficiency. It was also shown that expeller performance in terms of its IR improves with smaller hole size. Therefore, to improve an expeller suction flow capability, while maintaining its performance efficiency (i.e. IR), larger number of the same or smaller holes should be considered. © 2015 Elsevier B.V. All rights reserved.
Keywords: Expellers Gas pipeline blowdown Purging and evacuation Combustible gas
1. Introduction Expellers are commonly used in the gas pipeline industry to remove (expel) remaining gas in an isolated section of a pipeline after the section has been blown down to ambient pressure through the blowdown stacks at either end. Once this first step is completed (i.e. blowdown from line pressure to ambient pressure), expellers are mounted either on one side or two sides of the pipeline section, again on the respective blowdown stacks. (Fig. 1). When the expeller(s) are turned on either or both ends, air is drawn into the pipeline section at the work site (Bacon, 2000; Huang et al., 1999; Chen et al., June 1998; Villa et al., November 1999; McElligott et al., 1998; Parker, 1989; Pankratov et al., December 1987), and hence drive the combustible gas through the pipe toward the
* Corresponding author. NOVA Chemicals Centre for Applied Research, 2928 e 16 Street N.E., Calgary, Alberta, Canada T2E 7K7. E-mail address: [email protected] (K.K. Botros). http://dx.doi.org/10.1016/j.jngse.2015.02.025 1875-5100/© 2015 Elsevier B.V. All rights reserved.
expeller (Fig. 1). Compressed air (up to 1034 kPa-g or 150 psig) is used to drive expellers, which allows the operator to control the amount of airflow ingress through the opening at the work-site as well to balance the amount of airflow going in either direction depending on the location of opening along the pipeline section in relation to the two ends where the expellers are mounted. The drive air is fed to the expeller by a portable air compressor in the field. One or two compressors (depending on the air flow requirement) are typically used. No additional moving parts are employed by expellers, but they operate on the venturi principle where low pressure is created (induced) at the throat of the expeller by the compressed drive air flowing through a set of holes at the bottom of the device, as shown in the schematic of Fig. 2. The venturi concept is similar to that employed in subsonic ejectors (Sun and Eames, Jun 1995; Huang et al., 1999; Chen et al., June 1998; Villa et al., November 1999). The industry provided a guideline (American Gas Association, June 2001) for the appropriate selection of expellers and the appropriate practices of evacuating gas pipeline sections of different lengths and diameters.
K.K. Botros et al. / Journal of Natural Gas Science and Engineering 23 (2015) 458e463
and the time it takes to completely evacuate the pipeline section. The performance of expellers is given by the induction ratio, IR, which is defined as the ratio of the mass flow of the suction flow ( m_ s ) induced by the expeller and the drive flow of the compressed air or gas (m_ d ) (Botros et al., 2007a). The induction ratio of an expeller installed on a pipeline system is related to the overall equivalent flow resistance coefficient (Ke) of the pipeline, referenced to the dynamic head at the throat of the expeller, and is expressed by the following relationship (Botros et al., 2007a, 2007b):
Nomenclatures Ao hole area of drive air Aexp expeller throat area C1 constant C2 constant Cd orifice discharge coefficient D pipe internal diameter g acceleration of gravity I.R. induction mass ratio Ke equivalent resistance coefficient drive air mass flow rate m_ d induced (or suction) mass flow rate m_ s Po drive air pressure ðVpurge Þmin minimum purge velocity ra density of air rg density of gas
IR ¼
Valve: full bore or reduced bore ball valve, Or plug valve
Drive Air
Drive Air
459
Opening Location
Isolated Pipeline Section Fig. 1. Evacuation of an isolated pipeline section with two expellers at both ends.
m_ s ¼ C1 ðKe ÞC2 m_ d
where the two constants C1 and C2 define the performance characteristics of the specific expeller type and size. The values of these constants can be determined from actual testing of the specific expellers subjected to a varying flow resistance at inlet. The value of the constant C2 was previously found to be approximately equal to 0.5 (Botros et al., 2007a, 2007b). Clearly, the higher the value of the constant C1 the higher the induction ratio (stronger expeller), and hence the larger the suction flow for a given drive airflow. In recent years, the spacing between pipeline block valves became longer, which was primarily brought about by accessibility constraints and rough trains, among other factors. Clearly, when the length of a pipeline section increases, the resistance to the suction flow increases, the suction flow induced by the expeller decreases and hence the purge velocity decreases. Therefore, it is important to install the appropriate expeller size and use sufficient drive air pressure such that the purge velocity is maintained above a minimum value to prevent stratification, i.e. two counter-flowing layers of gas on top and air at the bottom (American Gas Association, June 2001). This minimum purge velocity, ðVpurge Þmin , is determined by the difference in gas and air densities and pipe internal diameter via (American Gas Association, June 2001):
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ! ffi u u ra rg Vpurge min ¼ t gD ra þ rg
Fig. 2. The venturi concept at the throat of an expeller.
The flow rate of air drawn into the pipeline section through the opening, which is driven by the blow off gas drawn by the expeller(s) (suction flow) is dependent on the drive air flow and the drive air pressure, as well as the flow resistance to the suction flow introduced by the pipeline wall friction, entrance losses at the bottom of the stack, and all other fittings along the suction flow path. A methodology has been developed in (Botros et al., 2007a, 2007b), which allows proper quantification of the effects of these resistive elements on the effectiveness of the expelling procedure
(1)
(2)
Commercially available expellers are currently limited in size and capability in permitting enough drive airflow, and hence the required suction flow according to Eq. (1) and the constraint of Eq. (2) may not be realized. Botros and Hawryluk (Hawryluk and Botros, 2008) showed that mounting two expellers in parallel at one location does not necessarily increase the induced flow. Therefore, there is a need to explore innovative and cost effective means to improve the capability of a given expeller to permit higher drive airflow within its geometrical and size limitation. This in turn should result in inducing higher suction flow (via Eq. (1)), even if the performance in term of the IR may drop somewhat. The present paper presents results of measurements conducted on a typical (commercially available) expeller which was slightly modified, again within its geometrical and size limitation, to permit increase in the drive airflow. This was achieved by increasing the number and size of the drive air holes around the collar of the expeller. Flow resistance imposed by a pipeline section attached to the expeller/stack assembly is assimilated experimentally by a restriction orifice at the other end of a plenum attached to the bottom side of the stack. Performance characteristics were determined in terms of drive airflow, drive air pressure, induced (suction) flow and the IR. Several modifications were made to the tested expeller and the relative increase in the suction flow achieved in relation to the original design was quantified in relation to the potential loss in the IR. Section 2 provides description of the experimental setup,
460
K.K. Botros et al. / Journal of Natural Gas Science and Engineering 23 (2015) 458e463
and results of the various tests on the AS IS vs. modifications of the expellers are given in Section 3. Section 4 provides a discussion on the results and presents an argument that increasing the suction flow of a given expeller is possible at a cost of sacrificing a bit on the induction ratio.
2. Experimental setup Fig. 3 shows a schematic of the test rig used in the present investigation, which has slightly different geometry than that used in the previous work by the authors (Botros et al., 2007a, 2007b). The tested expeller is a 150 mm nominal size mounted on approx. 2.0 m long (200 mm ID) stack. The other side of the stack is connected to a 6.2 m long (250 mm ID) plenum (Fig. 3). The end of the plenum is closed with standard end-cap with a certain size central open hole (orifice) e Fig. 4. Two end-caps with different size orifices (75 mm and 150 mm) were used to vary the flow resistance. The expeller is driven by one or two portable air compressors, each is capable of driving maximum flow of 0.14 St m3/s (300 SCFM) and maximum pressure of 1034 kPa-g (150 psig). Instrumentations, data acquisition, accuracy and uncertainty were reported in (Botros et al., 2007a, 2007b). It should be pointed out that in this experimental setup, the fluid medium that is being pulled by the expeller is strictly air. In an actual field setup (e.g. an isolated pipeline section between two mainline block valves), however, this medium would be initially natural gas, which would be driven out overtime with air ingress through the cut at the work location. Eventually at the end of the expelling process, the medium being pulled would become entirely air, which represents the worst-case scenario from expelling perspective. This is because pressure drop in air is higher than natural gas due to its relatively heavier density under the same mean flow velocity. The commercial expeller tested has a throat diameter of 152.4 mm. The drive air goes through 12 holes, each of 2.35 mm in diameter, hence the total sum of drive flow area is 52.02 mm2. This is labeled as the original design of the expeller in Table 1. Four modifications were then made to this expeller. First, the original holes were plugged, and 24 new holes were drilled at a smaller diameter of 1.5875 mm each. This modification is labeled Mod 1 and was done to show the effects of decreasing the hole diameter while (more or less) maintain the same total hole areas of the drive air. The second modification (Mod 2) entailed increasing the diameter of the 24 holes in Mod 1e2.2352 mm, which will result in increasing the total hole areas to 94.17 mm2. In Mod 3, 12 of the 24 holes in Mod 2 were further increased in diameter to 3.0 mm, while in Mod 4 all 24 holes were enlarged to 3.0 mm. Table 1 shows a summary of these four modifications, and relative increase/ decrease of the total drive air hole areas relative to the original design. Fig. 5 shows photos through the expeller diffuser section
showing the 12 holes for the original design and the 24 holes for Mod 3 and Mod 4.
3. Test results and discussion Two sets of tests were conducted on the original expeller design and the four modifications outlined in Table 1. The first set was conducted with the larger orifice (150 mm diameter) at the back of the plenum (cf. Figs. 3 and 4). The second set is a repeat of the first set but with the smaller orifice (75 mm) at the back of the plenum. These two different orifice sizes represent a vast difference in the flow resistance in the range of approx.1:16. It is the flow resistance Ke with the 75 mm orifice is approximately 16 times higher than that of the larger orifice. The main purpose of this is to ascertain whether the performance of the expeller with the four modifications is similar regardless of the flow resistance of the system. Fig. 6 shows the experimental results of the measured suction flow vs. drive airflow through the original design of the expeller, and comparison with the four modifications outlines in Table 1. Note that the drive airflow is higher for Mod 2, 3 and 4 due to the larger total drive air hole areas. Clearly, with large drive air hole areas, the drive airflow increases for the same drive air pressure, though not quite linearly due to variations in the hole discharge flow coefficient. Fig. 7 presents the same suction flow results but vs. the drive air pressure. Both Figs. 6 and 7 show that the increase in the drive airflow brought about by the increase in the total drive air hole areas in Mod 2, 3 and 4 resulted in an increase in the suction flow. In fact at the same drive pressure (e.g. 690 kPag), the suction flow increases from 0.558 m3/s for the original design case to 1.047 m3/s for Mod 4 case (i.e. 1.88 times) for the 150 mm orifice test configuration. The corresponding drive airflow for these two cases are 0.094 m3/s and 0.233 m3/s, respectively, i.e. at a ratio of 2.49. Obviously, the increase in the suction flow (the 1.88 times) is not proportional to the increase in the drive air flow (the 2.49 times). This translates to a loss in the induction ratio (IR) from 5.97 to 4.49, i.e. by a ratio of 0.75. The main notion from all of this is that, the suction flow can be increased with more and larger drive air holes at the collar of the expeller, but at the cost of a less proportionate reduction in the IR. The former is the 1.88 ratio, and the latter is the 1.33 (inverse of 0.75), hence the net effect is a gain. If the former ratio turned out to be the same as the latter, then obviously, there would be no merits in increasing the number and hole sizes of an expeller. The above trend is also observed with the smaller orifice size (75 mm) on the back of the plenum, as indicated by the set of results on both Figs. 6 and 7. At same drive pressure of 690 kPag, the suction flow is 0.128 m3/s for the original design and 0.285 m3/s for Mod 4, i.e. an increase of 2.230 times. The corresponding IR dropped from 2.25 for the original design to 1.22 for Mod 4, i.e. a ratio of 1.84. Again the net effect is positive. The other modifications Mod 2
Fig. 3. Schematic of the Experimental Setup (all dimensions are in ‘mm’).
K.K. Botros et al. / Journal of Natural Gas Science and Engineering 23 (2015) 458e463
461
Fig. 4. Photo of the test setup at TCPL's gas dynamic test Facility in Didsbury, Alberta, Canada.
Table 1 Number and sizes of drive air holes of the original design expeller and the subsequent four modifications aimed at increasing the drive air flow. 150 mm Expeller
# of holes
Hole diameter (inches)
Hole diameter (mm)
Total hole area, Ao (mm2)
No. of drive compressors
Ratio of holes area to original design
Original design Mod 1 Mod 2 Mod 3 Mod 4
12 24 24 24 24
0.0925 0.0625 0.088 (12 0.088) þ (12 0.118) 0.118
2.3500 1.5875 2.2352 (12 2.2352) þ (12 3.0) 3
52.05 47.50 94.17 136.87 169.65
1 1 1 2 2
1.00 0.91 1.81 2.63 3.26
and Mod 3 also show the same trend, in that the gains in the suction flow surpass the loss in the induction ratio, and hence overall improvement in the suction flow capability of the expeller (at the cost of efficiency in terms IR). Fig. 8 gives the overall trending of the IR as the suction flow increases of all of the tested cases. It clearly shows a loss in the IR as the suction flow increases with increase in the drive pressure and/or with large number and size of drive air holes. However, the slope of IR reduction is smaller than the slope of the increase in the suction flow (cf. Fig. 6), and hence there is a net gain overall. Table 2 provide the above comparison and expands it to more detail to include Mod 1, 2 and 3 for the example drive air pressure of 690 kPag considered above. The trend of Mod 3 follows that of Mod 4 in comparison to the original design. Examining the results of Mod1 and Mod 2 in Table 2, there appears to be additional important observation about these two modifications, both of which share the same trend of higher IR than that of the original design. Recall that Mod 1 and Mod 2 have smaller drive air hole sizes of 1.5875 mm and 2.2352 mm, respectively, than that of the original design (2.35 mm). This resulted in improvement in the expeller performance in terms of its IR. Obviously, Mod 2 has the
benefits of both improvements: increase in the suction flow (brought about by the increase in the total are hole area by 1.81 times, and a modest increase of its IR to 1.019 times (cf. Table 2, Mod 2 line). It is important to realize that evacuating the pipeline from the remaining combustible gases in a safe and timely manner is critical. Prolonged expelling time results in extending the overall pipeline shutdown period and hence unfavorable economic impact. Prolonged expelling time would also result in extending the work-site crew hours at a substantial cost. On the other hand, the cost of an additional one or two air compressors deployed to the expellers sites to increase the drive air flow (regardless of the loss in the IR) is less than the cost impact of a longer expelling time as mentioned above. 4. Conclusions 1. An expeller performance in terms of its capability of driving higher suction flows from a given flow resistance system can be improved by increasing the number and size of the drive air holes drilled along the circumference of the expeller throat. This
Fig. 5. Photo through the expeller diffuser showing the drive air holes of the original design and Mod 3 and Mod 4 modifications.
462
K.K. Botros et al. / Journal of Natural Gas Science and Engineering 23 (2015) 458e463
Fig. 6. Experimental results of suction flow vs. drive airflow through the tested expeller showing improvement of the suction flow with modifications Mod 2, Mod 3 and Mod 4.
Fig. 7. Experimental results of suction flow vs. drive air pressure through the tested expeller showing improvement of the suction flow with modifications Mod 2, Mod 3 and Mod 4.
in turn will permit higher drive air flow, which in practice, will require more than one air compressor if the drive medium is air. If the drive medium is natural gas drawn from the adjacent pipeline, then there should be enough drive flow capacity available for the expeller. 2. However, with increased drive flows, and thus achieving the desired increase in the suction flow capability of the expeller, the performance of the expeller in terms of its induction ratio (IR) deteriorates, but luckily not at the same rate as the suction
flow increases. Hence a cost effective means to improve the suction flow capability of an expeller is to drill more and larger size holes around its throat. The loss in the IR (which is efficiency related), however, is generally not a concern in practice, particularly when the economic benefit of evacuating the pipeline section in a timely and safe manner greatly outweighs any potential loss in the expeller IR efficiency. 3. The smaller the hole size for the drive air flow, the higher the IR. Therefore, to improve on an expeller suction flow capability,
K.K. Botros et al. / Journal of Natural Gas Science and Engineering 23 (2015) 458e463
463
Fig. 8. Experimental results of the induction ration of the test expeller showing deterioration in the induction ratio with modifications Mod 2, Mod 3 and Mod 4.
Table 2 Performance comparison of the original design expeller vs. the four modifications at drive air pressure of 690 kPag (100 psig) for the case of 150 mm orifce at the back of the plenum. Comparison at 690 kPag drive air pressure 150 mm Expeller
# of hole
Hole diameter (inches)
Hole diameter (mm)
Total hole area, Ao (mm2)
No. of drive compressors
Ratio of hole area to original design
Drive flow (St.m3/s)
Ratio of drive flow
Suction flow (St.m3/s)
Ratio of suction flow
IR
Ratio of IR to original design
Original design Mod 1 Mod 2 Mod 3
12 24 24 24
1 1 1 2
1.00 0.91 1.81 2.63
0.09 0.08 0.13 0.19
1.00 0.90 1.38 2.05
0.56 0.57 0.79 0.89
1.00 1.01 1.41 1.59
5.97 6.74 6.08 4.63
1.00 1.13 1.02 0.78
24
2.3500 1.5875 2.2352 (12 2.2352) þ (12 3.0) 3
52.05 47.50 94.17 136.87
Mod 4
0.0925 0.0625 0.088 (12 0.088) þ (12 0.118) 0.118
169.65
2
3.26
0.23
2.49
1.05
1.88
4.49
0.75
while maintaining its performance efficiency (i.e. IR), larger number of the same or smaller size holes should be considered. Acknowledgments The authors appreciate the discussion with Tom Thrall, Darrel Sayer, Dmitry Ryapolov and James Ferguson of TCPL during the course of this work, as well as the support of TCPL Technology Management; Thomas Robinson, Anthony Tse and Tracy Cairns during the course of this work. This paper is part of a research program sponsored by TransCanada Pipelines Limited, and permission to publish is gratefully acknowledged. References American Gas Association, June 2001. Purging Principle and Practice, Third ed. Catalog No. XK0101. Bacon, D., 2000. Isolation techniques for pipeline maintenance and decommissioning. In: 4th Int. IBC Global Conf., Onshore Pipelines, pp. 1e18. Botros, K.K., Geerligs, J., Sadoway, B., Watson, B., 2007. Performance of expellers in
evacuating gas pipelines e Part I: measurements and models. Int. J. Press. Vessel Pip. 84, 412e422. Botros, K.K., Geerligs, J., Sadoway, B., Watson, B., 2007. Performance of expellers in evacuating Gas pipelines e Part II: effects of plug valve and natural gas as a drive gas. Int. J. Press. Vessel Pip. 84, 423e429. Chen, S.L., Yen, J.Y., Huang, M.C., June 1998. An experimental investigation of ejector performance based upon different refrigerants. ASHRAE Trans. 104 (2), 153e160. Hawryluk, A., Botros, K.K., 2008. Effectiveness of evacuating combustible gases by two parallel expellers closely coupled at one end of a gas pipeline. Int. J. Press. Vessel Pip. 85, 860e865. Huang, B.J., Chang, J.M., Wang, J.M., Petrenko, V.A., 1999. A 1-D analysis of ejector performance. Int. J. Refrig. 22, 354e364. McElligott, J.A., Delanty, J., Delanty, B., 1998. Use of hot taps for gas pipelines can be expanded. In: ASME International Pipeline Conference, June 7-177. Pankratov, V.S., Sipershtein, B.I., December 1987. Determining the evacuation time for a gas pipeline. Gazov. Promst. 39e41. Parker, R.F., 1989. Get the air out, then test, then get the water out. In: 12th Annu. ASME Energy-Sources Technology Conf., Pipeline Eng. Symp., Houston, January 22e25. Sun, Da-Wen, Eames, I.W., Jun 1995. Recent developments in the design theories and applications of ejectors e a review. J. Inst. Energy 68 (475), 65e79. Villa, M., De Ghetto, G., Paone, F., Giacchetta, G., Bevilacqua, M., November 1999. Ejectors for boosting low-pressure oil wells. SPE Prod. Facil. 14 (4), 229e234.
Contents lists available at ScienceDirect
Journal of Natural Gas Science and Engineering journal homepage: www.elsevier.com/locate/jngse
Experimental investigation aiming at improving the suction flow capability of a gas expeller Kamal K. Botros a, *, John Geerligs a, Brad Watson b a b
NOVA Chemicals Centre for Applied Research, Calgary, Alberta, Canada TransCanada Pipelines Limited, Calgary, Alberta, Canada
a r t i c l e i n f o
a b s t r a c t
Article history: Received 14 January 2015 Received in revised form 18 February 2015 Accepted 19 February 2015 Available online 3 March 2015
An expeller performance has been evaluated in terms of its capability to induce higher suction flow for application to evacuate combustible gases from a blown down natural gas pipelines. The investigation involved a test rig and testing of a typical 150 mm (nominal size) expeller. This particular expeller has 12, 2.35 mm diameter holes, equally spaced around the throat circumference of the expeller. This was referenced to as the base (or original design). The aim of the present investigation is to improve the suction flow capability of this expeller by four modifications to both the number and/or sizes of these holes. The experimental results showed that the performance of the expeller in terms of its capability of driving higher suction flows for a given flow resistance system can be improved by increasing the number and sizes of the drive air holes which in turn permit higher drive air flow. However, with increased drive air flow, the performance of the expeller in terms of the induction ratio (IR) deteriorates, but luckily not at the same rate as the suction flow increases. Hence a cost effective means to improve the suction flow capability of an expeller is to drill more and larger size holes around its throat. The loss in the IR (which is efficiency related), however, is generally not a concern in practice when the economic benefit of evacuating the pipeline section in a timely and safe manner greatly overweigh any potential loss in the expeller IR efficiency. It was also shown that expeller performance in terms of its IR improves with smaller hole size. Therefore, to improve an expeller suction flow capability, while maintaining its performance efficiency (i.e. IR), larger number of the same or smaller holes should be considered. © 2015 Elsevier B.V. All rights reserved.
Keywords: Expellers Gas pipeline blowdown Purging and evacuation Combustible gas
1. Introduction Expellers are commonly used in the gas pipeline industry to remove (expel) remaining gas in an isolated section of a pipeline after the section has been blown down to ambient pressure through the blowdown stacks at either end. Once this first step is completed (i.e. blowdown from line pressure to ambient pressure), expellers are mounted either on one side or two sides of the pipeline section, again on the respective blowdown stacks. (Fig. 1). When the expeller(s) are turned on either or both ends, air is drawn into the pipeline section at the work site (Bacon, 2000; Huang et al., 1999; Chen et al., June 1998; Villa et al., November 1999; McElligott et al., 1998; Parker, 1989; Pankratov et al., December 1987), and hence drive the combustible gas through the pipe toward the
* Corresponding author. NOVA Chemicals Centre for Applied Research, 2928 e 16 Street N.E., Calgary, Alberta, Canada T2E 7K7. E-mail address: [email protected] (K.K. Botros). http://dx.doi.org/10.1016/j.jngse.2015.02.025 1875-5100/© 2015 Elsevier B.V. All rights reserved.
expeller (Fig. 1). Compressed air (up to 1034 kPa-g or 150 psig) is used to drive expellers, which allows the operator to control the amount of airflow ingress through the opening at the work-site as well to balance the amount of airflow going in either direction depending on the location of opening along the pipeline section in relation to the two ends where the expellers are mounted. The drive air is fed to the expeller by a portable air compressor in the field. One or two compressors (depending on the air flow requirement) are typically used. No additional moving parts are employed by expellers, but they operate on the venturi principle where low pressure is created (induced) at the throat of the expeller by the compressed drive air flowing through a set of holes at the bottom of the device, as shown in the schematic of Fig. 2. The venturi concept is similar to that employed in subsonic ejectors (Sun and Eames, Jun 1995; Huang et al., 1999; Chen et al., June 1998; Villa et al., November 1999). The industry provided a guideline (American Gas Association, June 2001) for the appropriate selection of expellers and the appropriate practices of evacuating gas pipeline sections of different lengths and diameters.
K.K. Botros et al. / Journal of Natural Gas Science and Engineering 23 (2015) 458e463
and the time it takes to completely evacuate the pipeline section. The performance of expellers is given by the induction ratio, IR, which is defined as the ratio of the mass flow of the suction flow ( m_ s ) induced by the expeller and the drive flow of the compressed air or gas (m_ d ) (Botros et al., 2007a). The induction ratio of an expeller installed on a pipeline system is related to the overall equivalent flow resistance coefficient (Ke) of the pipeline, referenced to the dynamic head at the throat of the expeller, and is expressed by the following relationship (Botros et al., 2007a, 2007b):
Nomenclatures Ao hole area of drive air Aexp expeller throat area C1 constant C2 constant Cd orifice discharge coefficient D pipe internal diameter g acceleration of gravity I.R. induction mass ratio Ke equivalent resistance coefficient drive air mass flow rate m_ d induced (or suction) mass flow rate m_ s Po drive air pressure ðVpurge Þmin minimum purge velocity ra density of air rg density of gas
IR ¼
Valve: full bore or reduced bore ball valve, Or plug valve
Drive Air
Drive Air
459
Opening Location
Isolated Pipeline Section Fig. 1. Evacuation of an isolated pipeline section with two expellers at both ends.
m_ s ¼ C1 ðKe ÞC2 m_ d
where the two constants C1 and C2 define the performance characteristics of the specific expeller type and size. The values of these constants can be determined from actual testing of the specific expellers subjected to a varying flow resistance at inlet. The value of the constant C2 was previously found to be approximately equal to 0.5 (Botros et al., 2007a, 2007b). Clearly, the higher the value of the constant C1 the higher the induction ratio (stronger expeller), and hence the larger the suction flow for a given drive airflow. In recent years, the spacing between pipeline block valves became longer, which was primarily brought about by accessibility constraints and rough trains, among other factors. Clearly, when the length of a pipeline section increases, the resistance to the suction flow increases, the suction flow induced by the expeller decreases and hence the purge velocity decreases. Therefore, it is important to install the appropriate expeller size and use sufficient drive air pressure such that the purge velocity is maintained above a minimum value to prevent stratification, i.e. two counter-flowing layers of gas on top and air at the bottom (American Gas Association, June 2001). This minimum purge velocity, ðVpurge Þmin , is determined by the difference in gas and air densities and pipe internal diameter via (American Gas Association, June 2001):
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ! ffi u u ra rg Vpurge min ¼ t gD ra þ rg
Fig. 2. The venturi concept at the throat of an expeller.
The flow rate of air drawn into the pipeline section through the opening, which is driven by the blow off gas drawn by the expeller(s) (suction flow) is dependent on the drive air flow and the drive air pressure, as well as the flow resistance to the suction flow introduced by the pipeline wall friction, entrance losses at the bottom of the stack, and all other fittings along the suction flow path. A methodology has been developed in (Botros et al., 2007a, 2007b), which allows proper quantification of the effects of these resistive elements on the effectiveness of the expelling procedure
(1)
(2)
Commercially available expellers are currently limited in size and capability in permitting enough drive airflow, and hence the required suction flow according to Eq. (1) and the constraint of Eq. (2) may not be realized. Botros and Hawryluk (Hawryluk and Botros, 2008) showed that mounting two expellers in parallel at one location does not necessarily increase the induced flow. Therefore, there is a need to explore innovative and cost effective means to improve the capability of a given expeller to permit higher drive airflow within its geometrical and size limitation. This in turn should result in inducing higher suction flow (via Eq. (1)), even if the performance in term of the IR may drop somewhat. The present paper presents results of measurements conducted on a typical (commercially available) expeller which was slightly modified, again within its geometrical and size limitation, to permit increase in the drive airflow. This was achieved by increasing the number and size of the drive air holes around the collar of the expeller. Flow resistance imposed by a pipeline section attached to the expeller/stack assembly is assimilated experimentally by a restriction orifice at the other end of a plenum attached to the bottom side of the stack. Performance characteristics were determined in terms of drive airflow, drive air pressure, induced (suction) flow and the IR. Several modifications were made to the tested expeller and the relative increase in the suction flow achieved in relation to the original design was quantified in relation to the potential loss in the IR. Section 2 provides description of the experimental setup,
460
K.K. Botros et al. / Journal of Natural Gas Science and Engineering 23 (2015) 458e463
and results of the various tests on the AS IS vs. modifications of the expellers are given in Section 3. Section 4 provides a discussion on the results and presents an argument that increasing the suction flow of a given expeller is possible at a cost of sacrificing a bit on the induction ratio.
2. Experimental setup Fig. 3 shows a schematic of the test rig used in the present investigation, which has slightly different geometry than that used in the previous work by the authors (Botros et al., 2007a, 2007b). The tested expeller is a 150 mm nominal size mounted on approx. 2.0 m long (200 mm ID) stack. The other side of the stack is connected to a 6.2 m long (250 mm ID) plenum (Fig. 3). The end of the plenum is closed with standard end-cap with a certain size central open hole (orifice) e Fig. 4. Two end-caps with different size orifices (75 mm and 150 mm) were used to vary the flow resistance. The expeller is driven by one or two portable air compressors, each is capable of driving maximum flow of 0.14 St m3/s (300 SCFM) and maximum pressure of 1034 kPa-g (150 psig). Instrumentations, data acquisition, accuracy and uncertainty were reported in (Botros et al., 2007a, 2007b). It should be pointed out that in this experimental setup, the fluid medium that is being pulled by the expeller is strictly air. In an actual field setup (e.g. an isolated pipeline section between two mainline block valves), however, this medium would be initially natural gas, which would be driven out overtime with air ingress through the cut at the work location. Eventually at the end of the expelling process, the medium being pulled would become entirely air, which represents the worst-case scenario from expelling perspective. This is because pressure drop in air is higher than natural gas due to its relatively heavier density under the same mean flow velocity. The commercial expeller tested has a throat diameter of 152.4 mm. The drive air goes through 12 holes, each of 2.35 mm in diameter, hence the total sum of drive flow area is 52.02 mm2. This is labeled as the original design of the expeller in Table 1. Four modifications were then made to this expeller. First, the original holes were plugged, and 24 new holes were drilled at a smaller diameter of 1.5875 mm each. This modification is labeled Mod 1 and was done to show the effects of decreasing the hole diameter while (more or less) maintain the same total hole areas of the drive air. The second modification (Mod 2) entailed increasing the diameter of the 24 holes in Mod 1e2.2352 mm, which will result in increasing the total hole areas to 94.17 mm2. In Mod 3, 12 of the 24 holes in Mod 2 were further increased in diameter to 3.0 mm, while in Mod 4 all 24 holes were enlarged to 3.0 mm. Table 1 shows a summary of these four modifications, and relative increase/ decrease of the total drive air hole areas relative to the original design. Fig. 5 shows photos through the expeller diffuser section
showing the 12 holes for the original design and the 24 holes for Mod 3 and Mod 4.
3. Test results and discussion Two sets of tests were conducted on the original expeller design and the four modifications outlined in Table 1. The first set was conducted with the larger orifice (150 mm diameter) at the back of the plenum (cf. Figs. 3 and 4). The second set is a repeat of the first set but with the smaller orifice (75 mm) at the back of the plenum. These two different orifice sizes represent a vast difference in the flow resistance in the range of approx.1:16. It is the flow resistance Ke with the 75 mm orifice is approximately 16 times higher than that of the larger orifice. The main purpose of this is to ascertain whether the performance of the expeller with the four modifications is similar regardless of the flow resistance of the system. Fig. 6 shows the experimental results of the measured suction flow vs. drive airflow through the original design of the expeller, and comparison with the four modifications outlines in Table 1. Note that the drive airflow is higher for Mod 2, 3 and 4 due to the larger total drive air hole areas. Clearly, with large drive air hole areas, the drive airflow increases for the same drive air pressure, though not quite linearly due to variations in the hole discharge flow coefficient. Fig. 7 presents the same suction flow results but vs. the drive air pressure. Both Figs. 6 and 7 show that the increase in the drive airflow brought about by the increase in the total drive air hole areas in Mod 2, 3 and 4 resulted in an increase in the suction flow. In fact at the same drive pressure (e.g. 690 kPag), the suction flow increases from 0.558 m3/s for the original design case to 1.047 m3/s for Mod 4 case (i.e. 1.88 times) for the 150 mm orifice test configuration. The corresponding drive airflow for these two cases are 0.094 m3/s and 0.233 m3/s, respectively, i.e. at a ratio of 2.49. Obviously, the increase in the suction flow (the 1.88 times) is not proportional to the increase in the drive air flow (the 2.49 times). This translates to a loss in the induction ratio (IR) from 5.97 to 4.49, i.e. by a ratio of 0.75. The main notion from all of this is that, the suction flow can be increased with more and larger drive air holes at the collar of the expeller, but at the cost of a less proportionate reduction in the IR. The former is the 1.88 ratio, and the latter is the 1.33 (inverse of 0.75), hence the net effect is a gain. If the former ratio turned out to be the same as the latter, then obviously, there would be no merits in increasing the number and hole sizes of an expeller. The above trend is also observed with the smaller orifice size (75 mm) on the back of the plenum, as indicated by the set of results on both Figs. 6 and 7. At same drive pressure of 690 kPag, the suction flow is 0.128 m3/s for the original design and 0.285 m3/s for Mod 4, i.e. an increase of 2.230 times. The corresponding IR dropped from 2.25 for the original design to 1.22 for Mod 4, i.e. a ratio of 1.84. Again the net effect is positive. The other modifications Mod 2
Fig. 3. Schematic of the Experimental Setup (all dimensions are in ‘mm’).
K.K. Botros et al. / Journal of Natural Gas Science and Engineering 23 (2015) 458e463
461
Fig. 4. Photo of the test setup at TCPL's gas dynamic test Facility in Didsbury, Alberta, Canada.
Table 1 Number and sizes of drive air holes of the original design expeller and the subsequent four modifications aimed at increasing the drive air flow. 150 mm Expeller
# of holes
Hole diameter (inches)
Hole diameter (mm)
Total hole area, Ao (mm2)
No. of drive compressors
Ratio of holes area to original design
Original design Mod 1 Mod 2 Mod 3 Mod 4
12 24 24 24 24
0.0925 0.0625 0.088 (12 0.088) þ (12 0.118) 0.118
2.3500 1.5875 2.2352 (12 2.2352) þ (12 3.0) 3
52.05 47.50 94.17 136.87 169.65
1 1 1 2 2
1.00 0.91 1.81 2.63 3.26
and Mod 3 also show the same trend, in that the gains in the suction flow surpass the loss in the induction ratio, and hence overall improvement in the suction flow capability of the expeller (at the cost of efficiency in terms IR). Fig. 8 gives the overall trending of the IR as the suction flow increases of all of the tested cases. It clearly shows a loss in the IR as the suction flow increases with increase in the drive pressure and/or with large number and size of drive air holes. However, the slope of IR reduction is smaller than the slope of the increase in the suction flow (cf. Fig. 6), and hence there is a net gain overall. Table 2 provide the above comparison and expands it to more detail to include Mod 1, 2 and 3 for the example drive air pressure of 690 kPag considered above. The trend of Mod 3 follows that of Mod 4 in comparison to the original design. Examining the results of Mod1 and Mod 2 in Table 2, there appears to be additional important observation about these two modifications, both of which share the same trend of higher IR than that of the original design. Recall that Mod 1 and Mod 2 have smaller drive air hole sizes of 1.5875 mm and 2.2352 mm, respectively, than that of the original design (2.35 mm). This resulted in improvement in the expeller performance in terms of its IR. Obviously, Mod 2 has the
benefits of both improvements: increase in the suction flow (brought about by the increase in the total are hole area by 1.81 times, and a modest increase of its IR to 1.019 times (cf. Table 2, Mod 2 line). It is important to realize that evacuating the pipeline from the remaining combustible gases in a safe and timely manner is critical. Prolonged expelling time results in extending the overall pipeline shutdown period and hence unfavorable economic impact. Prolonged expelling time would also result in extending the work-site crew hours at a substantial cost. On the other hand, the cost of an additional one or two air compressors deployed to the expellers sites to increase the drive air flow (regardless of the loss in the IR) is less than the cost impact of a longer expelling time as mentioned above. 4. Conclusions 1. An expeller performance in terms of its capability of driving higher suction flows from a given flow resistance system can be improved by increasing the number and size of the drive air holes drilled along the circumference of the expeller throat. This
Fig. 5. Photo through the expeller diffuser showing the drive air holes of the original design and Mod 3 and Mod 4 modifications.
462
K.K. Botros et al. / Journal of Natural Gas Science and Engineering 23 (2015) 458e463
Fig. 6. Experimental results of suction flow vs. drive airflow through the tested expeller showing improvement of the suction flow with modifications Mod 2, Mod 3 and Mod 4.
Fig. 7. Experimental results of suction flow vs. drive air pressure through the tested expeller showing improvement of the suction flow with modifications Mod 2, Mod 3 and Mod 4.
in turn will permit higher drive air flow, which in practice, will require more than one air compressor if the drive medium is air. If the drive medium is natural gas drawn from the adjacent pipeline, then there should be enough drive flow capacity available for the expeller. 2. However, with increased drive flows, and thus achieving the desired increase in the suction flow capability of the expeller, the performance of the expeller in terms of its induction ratio (IR) deteriorates, but luckily not at the same rate as the suction
flow increases. Hence a cost effective means to improve the suction flow capability of an expeller is to drill more and larger size holes around its throat. The loss in the IR (which is efficiency related), however, is generally not a concern in practice, particularly when the economic benefit of evacuating the pipeline section in a timely and safe manner greatly outweighs any potential loss in the expeller IR efficiency. 3. The smaller the hole size for the drive air flow, the higher the IR. Therefore, to improve on an expeller suction flow capability,
K.K. Botros et al. / Journal of Natural Gas Science and Engineering 23 (2015) 458e463
463
Fig. 8. Experimental results of the induction ration of the test expeller showing deterioration in the induction ratio with modifications Mod 2, Mod 3 and Mod 4.
Table 2 Performance comparison of the original design expeller vs. the four modifications at drive air pressure of 690 kPag (100 psig) for the case of 150 mm orifce at the back of the plenum. Comparison at 690 kPag drive air pressure 150 mm Expeller
# of hole
Hole diameter (inches)
Hole diameter (mm)
Total hole area, Ao (mm2)
No. of drive compressors
Ratio of hole area to original design
Drive flow (St.m3/s)
Ratio of drive flow
Suction flow (St.m3/s)
Ratio of suction flow
IR
Ratio of IR to original design
Original design Mod 1 Mod 2 Mod 3
12 24 24 24
1 1 1 2
1.00 0.91 1.81 2.63
0.09 0.08 0.13 0.19
1.00 0.90 1.38 2.05
0.56 0.57 0.79 0.89
1.00 1.01 1.41 1.59
5.97 6.74 6.08 4.63
1.00 1.13 1.02 0.78
24
2.3500 1.5875 2.2352 (12 2.2352) þ (12 3.0) 3
52.05 47.50 94.17 136.87
Mod 4
0.0925 0.0625 0.088 (12 0.088) þ (12 0.118) 0.118
169.65
2
3.26
0.23
2.49
1.05
1.88
4.49
0.75
while maintaining its performance efficiency (i.e. IR), larger number of the same or smaller size holes should be considered. Acknowledgments The authors appreciate the discussion with Tom Thrall, Darrel Sayer, Dmitry Ryapolov and James Ferguson of TCPL during the course of this work, as well as the support of TCPL Technology Management; Thomas Robinson, Anthony Tse and Tracy Cairns during the course of this work. This paper is part of a research program sponsored by TransCanada Pipelines Limited, and permission to publish is gratefully acknowledged. References American Gas Association, June 2001. Purging Principle and Practice, Third ed. Catalog No. XK0101. Bacon, D., 2000. Isolation techniques for pipeline maintenance and decommissioning. In: 4th Int. IBC Global Conf., Onshore Pipelines, pp. 1e18. Botros, K.K., Geerligs, J., Sadoway, B., Watson, B., 2007. Performance of expellers in
evacuating gas pipelines e Part I: measurements and models. Int. J. Press. Vessel Pip. 84, 412e422. Botros, K.K., Geerligs, J., Sadoway, B., Watson, B., 2007. Performance of expellers in evacuating Gas pipelines e Part II: effects of plug valve and natural gas as a drive gas. Int. J. Press. Vessel Pip. 84, 423e429. Chen, S.L., Yen, J.Y., Huang, M.C., June 1998. An experimental investigation of ejector performance based upon different refrigerants. ASHRAE Trans. 104 (2), 153e160. Hawryluk, A., Botros, K.K., 2008. Effectiveness of evacuating combustible gases by two parallel expellers closely coupled at one end of a gas pipeline. Int. J. Press. Vessel Pip. 85, 860e865. Huang, B.J., Chang, J.M., Wang, J.M., Petrenko, V.A., 1999. A 1-D analysis of ejector performance. Int. J. Refrig. 22, 354e364. McElligott, J.A., Delanty, J., Delanty, B., 1998. Use of hot taps for gas pipelines can be expanded. In: ASME International Pipeline Conference, June 7-177. Pankratov, V.S., Sipershtein, B.I., December 1987. Determining the evacuation time for a gas pipeline. Gazov. Promst. 39e41. Parker, R.F., 1989. Get the air out, then test, then get the water out. In: 12th Annu. ASME Energy-Sources Technology Conf., Pipeline Eng. Symp., Houston, January 22e25. Sun, Da-Wen, Eames, I.W., Jun 1995. Recent developments in the design theories and applications of ejectors e a review. J. Inst. Energy 68 (475), 65e79. Villa, M., De Ghetto, G., Paone, F., Giacchetta, G., Bevilacqua, M., November 1999. Ejectors for boosting low-pressure oil wells. SPE Prod. Facil. 14 (4), 229e234.