CHOPS operational and monitoring issues
While typical production operations seek to prevent sand production, cold heavy oil production with sand (CHOPS) operations use sand production to increase overall productivity. This difference can create operational issues throughout the life of a CHOPS well. It has implications for monitoring strategies as well.
- 1 Well completion practices
- 2 Production-decline mechanisms
- 3 Workover strategies
- 4 Monitoring produced fluids and well behavior
- 5 Fieldwide monitoring approaches
- 6 References
- 7 Noteworthy papers in OnePetro
- 8 External links
- 9 See also
- 10 Page champions
- 11 Category
Well completion practices
Perforating a CHOPS well
To initiate sand influx, a cased well is perforated with large-diameter ports, usually of 23 to 28 mm diameter, fully phased, and spaced at 26 or 39 charges per meter. More densely spaced charges have not proved to give better results or service, but less densely spaced charges (13 per meter) give poorer results. More densely spaced charges may eliminate reperforating as a future stimulation choice because full casing rupture is likely to take place. In thin intervals (<6 m), the entire interval is perforated. In thicker intervals (> 10 m), a 6- to 8-m zone is perforated. Shallow-penetrating perforations with large explosive charges shock and disrupt an annulus of sand around each perforation channel and, therefore, around the entire casing because these damaged zones likely overlap. Fig. 1 illustrates how aggressive perforation damages formation but helps initiate CHOPS. Shaped charges that leave a negligible amount of metallic debris in the formation are used to avoid impairing the elastomer in the PC pump when debris re-enters the well.
The optimum perforation-placement strategy is debatable. The perforated zone is either the bottom of the producing interval or the subzone with the best kh/μ value. Insufficient data exist to claim one approach is superior to the other. The author favors perforating in the lower part of the reservoir, retaining the flexibility to add perforations later higher in the interval. However, it is probably worthwhile to avoid perforating gravel zones with D50 > 2 to 3 mm because gravel is more likely to block perforation ports and damage the pump. In many cases, particularly in channels, gravel zones are at the base of the interval. The lowermost perforations should be placed above these zones.
Initiating production and pumping in CHOPS wells
The large sand influx in CHOPS wells increases well rates, but it carries the risk that the pump may plug and a workover may be required. Therefore, bringing a well onto production is a gradual process.
The pump is started while liquid, called load fluid, and usually lighter oil is introduced in the annulus. When the system is flowing freely, the load fluid rate is diminished to increase drawdown, and sand should begin to enter the wellbore. Drawdown and load fluid rate can be balanced to keep the pump operating effectively, but eventually load fluid input is stopped (within a day or two). If well capacity is sufficient, the pump is operated at maximum speed for the torque output, which is controlled at the drive head. If pump capacity exceeds the well capacity to deliver slurry, a lower speed is used.
CHOPS wells are maintained in an aggressively drawn-down condition, which increases the effectiveness of the foamy-oil mechanism in destabilizing sand and maintaining free flow into the perforations (low backpressure encourages gas and formation expansion). If BHP data are continuously available, they can be used to control the pump speed to maintain a 15- to 20-m annulus fluid level. Otherwise, regular fluid-level measurements are taken acoustically to optimize well performance.
Gas will evolve from the annulus because of gas breaking out of the slurry as it enters the wellbore and flows down past the stator housing. In Canada, this gas is collected and used to run pump motors and to heat the oil storage tanks on the production site.
Progressing cavity (PC) pumps
PC pumps currently are widely preferred over other pumps. Starting from poor life spans in the early 1980s, they have evolved into highly reliable, versatile devices capable of pumping slurries with high sand content for 15 to 20 months. Rapid technology advances are occurring, including the use of "sloppy-fit" pumps in heavy oil, two-stage pumps, helicoidal stator housings, compact surface hydraulic drives, and so on.
A few guidelines for PC pump use that have evolved from practice are included here.
- Wells should be drilled with a 20- to 30-m-deep rathole below the production zone to hold any large slugs of sand that enter during the production initiation phase.
- Pumps (7 to 70 m3/d/100 rev/min) should be sized appropriately for the expected volumes and depth.
- A 60- to 100-cm-long tailpipe with large vertical slots (10 mm wide, 100 mm long), open bottom, and welded horizontal tag bar (to prevent rotor drop in case of a failure) is attached to the end of the stator.
- The stator is installed in the wellbore with 3½-in. tubing with its base landed 1 m below the lowermost perforation.
- The antitorque anchor tool may be installed above or below the stator, depending on factors such as desire to maximize annulus gas recovery (tool below) or fear of plugging (tool above).
- Rotors are chromium plated or treated with boron to resist sand erosion and prolong life.
- During production, the end of the rotor should extend below the bottom of the stator by 10 to 50 cm so that it rotates in the upper part of the tailpipe.
- 25-mm ordinary continuous-drive rods or conventional sucker rods are used to drive the rotor. Various hard facings and antiwear devices are installed on the rods or the tubing to reduce tubing wear. (The relative value of different approaches remains to be assessed.)
- The pump must never be allowed to run dry through excessive formation gas throughput or because of excessive drawdown. The annulus must not be shut in so that gas buildup occurs.
- The rotor-seating level and the tubing and stator orientation must be changed at regular intervals to avoid wearing through the tubing.
Other lifting approaches
Reciprocating pumps are limited by slow rod-fall velocities. Gas lift is impractical, downhole jet pumps have been researched but never installed, and conventional electrosubmersible pumps still cannot handle sand.
If it is necessary to clean a well of a great deal of sand or if the sand cut is extremely high, a continuous sand-extraction pump based on stroking the tubing has been developed that can move sand at 60% porosity. PC pumps with rotating tubing that eliminate the rod string are under development, although casing wear will still be an issue. Downhole hydraulic drive for the rotor of the pump is available, and downhole electrical drives for PC pumps are being developed. For general use, the PC pump will likely remain the dominant device in CHOPS, but the advent of different drive systems will help improve lifting operations.
Pump manufacturing companies keep detailed records of reasons for pump failures. Reasons for pump failure include elastomer failure through ripping because of the intake of a piece of metal or a large pebble, elastomer embrittlement and failure if run dry, excessive rotor or stator wear, and torque off of the drive rods because of a sudden slug of sand. Additional reasons for pump failure are wear-through of the tubing, release of the no-turn anchor, and a failure of the surface drive system that allows sand to settle on the pump, which prevents startup.
In addition to mechanical problems, CHOPS wells may suffer production declines that are sudden or gradual. Careful data gathering for each well over time is needed for correct diagnosis.
For reasons that are poorly understood but probably related to stress-induced collapse and liquefaction of a region, a well that has been producing at a low sand cut may suddenly experience a massive, temporary influx of sand. Annulus blockage (between the stator and the casing) or pump blockage can occur. If the well begins to evidence a high water cut with a high sand rate, the lifting capacity may be seriously impaired. This allows sand to pile up in the tubing, causing rods to stop rotating or simply to plug upward flow. If power is removed suddenly from the surface equipment, sand settles onto the pump or slow sand influx into perforations blocks the flow path.
A frequent production-decline mechanism is the gradual plugging of perforations by sand, which is most common with small perforation ports and large sand grains. The drop in production occurs gradually until just a few perforations are producing. Production can cease suddenly if a sand slug is generated.
Some wells cease production not by perforation plugging but by near-wellbore sand recompaction. Perhaps the decline of solution-gas bubble drive allows the sand to stop moving and recompact. Compaction also may involve a segregation process: larger grains settle to the bottom of the liquefied zone, gradually blocking more perforations.
Gas tends to break out near the wellbore, filling the top of the liquefied zone and building a local gas cap that grows until it intersects the uppermost perforations (perhaps aided through coning) and eliminates oil production. Gas also can ruin the pump through overheating.
Remote reservoir processes
Lateral water coning has destroyed productivity prematurely in some CHOPS reservoirs. On the other hand, some fields have produced for more than 12 years with less than 20% water cut. In a part of Alberta's Lindbergh field located up a shallow dip from an active water zone, wells progressively watered out updip, indicating that the water source was a remote (800 to 2000 m distant) water zone. High drawdowns in 10,000 cp oil promote coning and, because conventional water shut-off technologies appear useless in CHOPS wells, solutions to coning are difficult.
Highly viscous oils that have resided in place for millions of years may be truly immobile under low gradients. In other words, the highly polar molecules (asphaltenes and resins) have become structured so that the substance has a small yield point that must be overcome. Encountering interwell virgin pressures in zones surrounded by long-term producing wells confirms the existence of a yield point, although the specific mechanics remain conjectural. As the disturbed zone around a CHOPS well grows and sand rates diminish, it is possible to "disconnect" the virgin far-field pressure regime from the sand-producing zone. Then, well productivity drops as the interior zone is depleted. Once this disconnect happens, the only solution is to break through the barrier.
Part of the reason for disconnection is that stable sand zones can develop in the interwell regions, similar to pillars in mines (Figs. 2 and 3). Because of low gradients and a fluid yield point, the sand can no longer be destabilized by gas exsolution and overburden stresses, leading to slow cessation of sand and fluid influx. In some fields (e.g., Lone Rock, Saskatchewan), wellbore pressures less than 1.0 MPa after 30 years shut-in indicate a lack of flow communication with the far field. However, many wells in that field have been rehabilitated successfully through aggressive workovers that "reconnected" with extant far-field pressures and likely destabilized the interwell stable regions so that gravity-induced sand drive could once again be generated.
Finally, general pressure depletion through depletion of the solution-gas drive causes irreversible production declines; however, whether gas can actually flow from a distance in an intact heavy-oil sandstone is open to question. Some fields have produced for more than a decade with constant gas/oil ratio (GOR), indicating a lack of drainage beyond the gas induction and damaged zone.
Given the reasons for production cessation, reinitiation or maintenance of sand influx is fundamental to all workover approaches. Without sand influx, oil rates will be uneconomical. Also, because of low fracture gradients in the well, no fluids will return to surface without artificial lift.
While the pump is running, fluids are introduced down the annulus, perhaps quite aggressively, to perturb the near-wellbore region, to help the pump move fluids, and to introduce chemicals into the wellbore region. Lighter oils (cleaned dead oil), perhaps of 17 to 20°API, can be heated to 70 to 90°C and pumped rapidly into the annulus. Alternatively, clean oil from the site stock tank (usually at 60 to 80°C because of tank heating and, therefore, lower viscosity) can be cycled back to the annulus. Even large slugs of water (approximately 5 to 10 m3), perhaps with chemicals to reduce capillary effects, can be flushed down the annulus while the pump is moving. These methods do not require a shutdown and are effective for near-wellbore problems only.
To avoid a complete workover, a winch truck can withdraw the rotor, and similar tactics can be used through the stator. In this case, as long as the tubing is flushed of sand, the treatment can be allowed to soak in the reservoir before the rotor is replaced and production reinitiated.
Well cleanout and perforation flushing
During a pump change out or tubing withdrawal, sand is cleaned from the well, and chemical soaks or other treatments may be used. Sand is removed with mechanical bailers, with pump-to-surface units (the well cannot give returns to surface), or with light foam-based workover fluids. In all cases, sand is removed to a depth of several meters below the perforations. This may require removing a great deal of sand because perforations will continue to produce sand, particularly when a mechanical wireline bailer is rapidly returned to surface, swabbing the wellbore. Perforations are flushed as thoroughly as possible, and chemical soaks are common at this juncture.
Reperforation and rocket propellant stimulation
Reperforation is used commonly to provide larger ports in old wells, to reduce sand inflow restrictions around the wellbore, and to perturb the wellbore environment and break down stable-sand zones. It is quite effective but, at most, two reperforations are feasible before the casing is ruptured. The beneficial effect likely extends no more than 1 to 2 m from the wellbore.
Ignition of small rocket propellant charges downhole has the benefit of blowing open all the perforations and shocking the zone around the wellbore with a large sudden outward surge. The shock probably perturbs a zone of approximately 4 to 6 m radius around the wellbore.
Pressure-pulse workovers, pulsed chemical placements
Aggressive pulsing to perturb the wellbore region was introduced in 1998. The liquid in a closed chamber (V ~ 200 L) at hole bottom, sealed from the annulus, is expelled suddenly through the perforations into the formation. The pressure pulse generated has a lower rise time and amplitude than with rocket propellant, but the impulse is repeated, perhaps 500 to 1,000 times in a 5-to 24-hour period, so that as much as 50 to 100 MJ of energy can be introduced. (Reperforation may involve 3 to 4 MJ, but at a high rate.) Beneficial effects appear to accumulate as more perforations are opened and the region near the wellbore is resaturated. The large energy input in each stroke has a cumulative effect, and an increase in pressure in the near-wellbore region typically reduces the gas saturation and allows the far-field effect to become more substantial. Measurements in offset wells (300 m) show that distant increases in flow rate may be triggered, showing that the effect propagates far beyond the wellbore. Pressure pulsing is the only method that can affect the interwell region by destabilizing distant zones and reconnecting the well with extant far-field pressures. It has been highly successful in initiating sanding in new wells that did not respond to standard sand-flow initiation procedures.
Aggressive pulsing, combined with the outward propagating wave of porosity dilation generated in each stroke, helps overcome problems of viscous fingering and preferential flow through high permeability streaks. This characteristic has resulted in new chemical-placement methods. Pulsing tool recharge from uphole (chemical) or from the formation can be varied to control concentrations of treatment fluids. The rapid downstroke forces fluids aggressively through the perforations, promoting full mixing with reservoir fluids. Fingering is suppressed by pulse placement; therefore, conformance is improved. Achieving good conformance during placement is particularly important in heavy oils because of slow diffusion rates.
Surface facilities and transportation
Because of the cold Canadian climate, the high oil viscosity and slow production rates, and the water cut in produced fluids, underground flowlines are used only on multiple well pads or if a well is adjacent to a battery or central facility. Tanker truck or load-haul-dump units transport produced streams (sand, oil, and water). Regional batteries collect and clean the oil and, except for one heated pipeline in Alberta, diluent (up to 15% naptha) is used for pipelining. Smaller-diameter diluent-return pipelines may be installed from upgraders to regional batteries.
Monitoring produced fluids and well behavior
To track and understand well behavior adequately, data that are more precise than the averaged data from stock-tank measurements is needed.
Sand content can be variable on any time scale. To examine this chaotic behavior, 500-mL flowline samples are taken. To obtain an averaged value, a 20-L pail is used, and 500-mL samples from the same well are added over time until approximately 5 L are available (10 to 15 samples). The gas is allowed to evolve before water, oil, and sand contents are determined. Standard oilfield methods are used (dilution with solvents, centrifuging, etc.) for solids and water contents.
Determining gas content is more difficult because gas is produced through the annulus as well as the tubing. Gas content measurements on flowline samples are straightforward: vacuum bottles are used to collect samples that are sent to a laboratory for analysis. The low-pressure gas exiting through the annulus can be metered (although this is still rare), and the two values corrected and combined to give an average gas content. In the absence of reliable gas information from which solubility can be calculated, a value for methane of 0.20 v/v/atm has been found to be widely consistent for heavy oil.
Long-term averages for sand, water, and oil volumes can be estimated from the stock-tank levels and transportation information. However, care must be taken in sand-volume estimation because it is easy to be in error by 10 to 15%, particularly if thick emulsions are involved. Installation of a bottomhole pressure (BHP) gauge on the tubing is advised, as well as continuous monitoring of both annulus pressure and pressure at the top of the pump. These data streams can be used directly in computer programs that optimize pumping, maintaining the fluid annulus at the desired level. Operating data for the pump (rev/min, torque) are used along with BHP data to ensure optimal pump operation.
Fieldwide monitoring approaches
Little full-field monitoring of CHOPS has occurred because of the low profit margins. However, the incentive to collect data for infill drilling and reservoir management is increasing, whereas costs for collecting data are decreasing.
Active seismic methods (3D seismic imaging), used by PanCanadian Petroleum, have identified large zones of low seismic velocity and large attenuation (Fig. 4), as well as interwell areas where depletion has not occurred. Passive seismic monitoring involves listening to seismic emissions generated by production processes. Because geophones are located in the reservoir, wave travel distances are short, and low-magnitude events are easily registered. This method has been used to track fireflood fronts and, more recently, to monitor activity in the reservoir during pressure pulsing in an excitation well. In the latter case, an increasing incidence of interwell shear events indicated that the pulsing had the effect of destabilizing the interwell region, allowing access to undepleted pressures. Microseismic monitoring has promise as a method of helping field management. Well testing is not a useful practice in CHOPS fields: interference phenomena, rate effects, etc. cannot be analyzed with existing theories.
- Kudu Industries Inc. Website, Kudu Industries Inc., Calgary, http://www.kudupump.com/.
- Weatherford. Artifical Lift. www.weatherford/com/divisions/artificiallift/pcp
- orlac Oil Production Solutions, Corlac, Lloydminster, Alberta, Canada, http://www.corlac.com/
- Kirby Hayes Inc. Website, Lloydminster, Alberta, Canada, http://www.kirbyhayes.com/.
- Dusseault, M.B. et al. 2000. Workover Strategies in CHOP Wells. Proc., CIM Petroleum Society 51st Annual Technical Meeting, Calgary, paper 2000-69.
- Dusseault, M.B., Spanos, T.J.T., and Davidson, B.C. 1999. A New Workover Tool—Applications in CHOP Wells. Proc., CIM Petroleum Society 50th Annual Tech. Meeting, Calgary.
- Dusseault, M.B. et al. 2001. Rehabilitating Heavy Oil Wells Using Pulsing Workovers to Place Treatment Chemicals. Proc., CIM Petroleum Society 52nd Annual Technical Meeting, Calgary, paper 2001-57.
- Nyland, E. and Dusseault, M.B. 1983. Fireflood Microseismic Monitoring: Results and Potential for Process Control. J. Cdn Pet Tech 22 (2): 62.
Noteworthy papers in OnePetro
Dusseault, M., Hayes, K., Kremer, M., & Wallin, C. 2004. Workover Strategies in CHOPS Wells. Petroleum Society of Canada. http://dx.doi.org/10.2118/04-09-GE
Young, J. P., Mathews, W. L., & Hulm, E. 2010. Alaskan Heavy Oil: First CHOPS at a vast, untapped arctic resource. Society of Petroleum Engineers. http://dx.doi.org/10.2118/133592-MS
Dusseault, M., Gall, C., Shand, D., Davidson, B., & Hayes, K. 2001. Rehabilitating Heavy Oil Wells Using Pulsing Workovers to Place Treatment Chemicals. Petroleum Society of Canada. http://dx.doi.org/10.2118/2001-057
Fernandez, D., Vidal, J. M., & Festini, D. E. 2012. Decision-Making Breakthrough Technology. Society of Petroleum Engineers. http://dx.doi.org/10.2118/152524-MS
Use this section to provide links to relevant material on websites other than PetroWiki and OnePetro