Tommy E. Taylor
Fasken Oil & Ranch Interests
Midland, Tex.
Lease operating expenses decreased by 82% after a sucker-rod pump replaced a jet pump in a deep, 12,500 ft West Texas well.
In designing the rod pumping system, a simulation program (wave equation) quantified possible ranges of equipment loading, rod loading, and plunger over-travel.
After 4 1/2 years of operating experience, some of the important aspects learned from the well are that:
- Commercial computer programs can accurately predict sucker-rod pumping performance in a deep well.
- Fiber glass sucker rods will significantly overtravel a pump plunger in a deep well.
- Conservative design and proper installation of a rod pumping system will minimize operating expenses and ultimately increase recoverable reserves.
WELL HISTORY
Barbara Fasken's Fee C is a one well lease located about 12 miles east of Andrews, Tex. Drilled in 1954 the 12,585 ft well was completed as a flowing oil well (Fig. 1) in the Magutex (Devonian) field.
In 1967, the well ceased flowing naturally and a sucker-rod pumping system was installed. The system included a 1.25-in. insert pump set at 8,500 ft, an API 76 rod-string design, and a size 320 pumping unit.
During the 22 months on pump, the well was pulled seven times: once to add a 4,000 ft segment of tail pipe, twice for pump repairs, and four times for rod failures. Each failure in the rod string was either a broken coupling, or a broken pin.
In 1969, an hydraulic reciprocating pump was installed to increase production to a top allowable of 130 bo/d. The reciprocating pump had to be pulled to surface and repaired frequently because of iron sulfide buildup in the engine end of the pump.
Two production casing leaks across the Grayburg interval (5,000 ft) had been squeezed prior to this time. The iron sulfide may have originated from this source.
To eliminate pump plugging problems, a jet pump was installed in 1986. The operating cost with the jet pump, however, was considerably higher than with the reciprocating pump. The increase was mainly due to the additional power oil required to operate the jet pump.
More power oil equates to higher electrical power costs.
The jet system did pump at the desired production rate of 50 bo/d and 10 bw/d, but as the well began to pump-off, the jet pump began cavitating.
Therefore, in 1988 a sucker rod system was installed to pump the well from 12,500 ft. The system has now operated for about 4.5 years without a single failure.
PUMPING EFFICIENCY
A comparison of the overall pumping efficiencies of two different artificial lift systems will indicate which system is more efficient in lifting a given amount of fluid from a given depth.1
Overall efficiency from measured data is the ratio of the useful pumping system output, Ho, divided by the input horsepower to the motor and multiplied by 100. The pumping efficiency at the pump discharge for the sucker-rod pumping system is shown in the example calculation box.
The pumping efficiency comparison indicates that, for this well, rod pumping is 3 1/2 times more efficient than jet pumping.
DESIGN CONSIDERATIONS
The sucker-rod pumping system was designed by using a commercially available rod-pumping simulation program (wave equation).2 Most of the input data for the predictive program are known quantities such as pumping unit linkages, sucker-rod properties, pump size, and tubing depth.
Other quantities such as downhole friction, pump intake pressure, and pump fill have to be estimated. Table 1 lists some of the important input data, predicted results, and measured data.
SUCKER RODS
The basic criteria for designing the rod string must be the rod stresses and the fatigue endurance of the rod string.3 Early in the design phase, a string was selected that consisted of 50% fiber glass 1.25-in. rods and 50% steel 7/8 in. and 3/4-in. rods.
Fiber glass rods were chosen because or their light-weight and tensile strength that is sufficient to handle the required rod loading. Fiber glass rods are also more elastic than steel rods, and therefore provide significant overtravel of the pump plunger.
This design results in a lightweight rod string that overtraveled the pump plunger, and the rod loadings fell below 100% loading range (service factor = 1.0).
Initially, an API 86 rod taper was selected (1-in., 7/8-in., and 3/4-in. taper). However, increasing the fiber glass rod diameter from 1 to 1.25 in. decreased the fiber glass rod loading from 66 to 53%, and the pump plunger travel increased from 156 to 196 in. at pumped-off conditions.
This is important because a decrease in fiber glass rod loading increases the expected cycles to first failure (Figs. 2-4).
Fig. 2 (API Goodman diagram) and Fig. 3 (stress range diagram) are used in the predictive program to calculate rod loadings for each rod size.
Fig. 4 is used for estimating cycles to first failure for the fiber glass rods. Note the significant increase in cycles to first failure below fiber glass rod loadings of 85%.
FRICTION
Several predictive runs were made by varying individual "critical" input parameters. This ensured that if any given actual load was higher than estimated, the pumping system would still perform under acceptable equipment loading to pump the desired amount of fluid.
The input downhole friction is one quantity that has to be estimated. This is a "critical" parameter because a severely deviated well bore could prevent successful rod pumping.
A deviation survey of this well indicated no severe departure from vertical. The maximum hole deviation was less than 3.
For normal hole deviation, a rule of thumb of 10% times the well depth can be used to estimate the downhole friction (0.10 x 12,500 ft = 1,250 lb or about 1,300 lb).
A predictive run at twice this friction indicated that the well could still be rod pumped, and the equipment would not be overloaded.
GAS ANCHOR
Above the perforations, a permanent packer was installed to form an artificial sump for separating produced liquids from produced gas.
This was done because pumping below the perforations was not feasible due to the plug-back TD being only 9 ft below the bottom of the perforations.
Fig. 5 shows the downhole separator arranged in a parallel packer configuration.
One drawback for using this type of separator is the possibility of solids, scale, etc. settling around the top of the packer.
This settling could create problems in releasing the seal divider during tubing retrieval. Also, fishing operations could be difficult in the section where the 1-in. tubing is banded to the 2.375-in. tubing.
This design, however, is a very effective gas separator because of the relatively low downward velocity of the liquids.4 The low velocity allows the gas to breakout of the liquid and rise up the tubing-casing annulus.
The packer configuration also provides a tubing anchoring device. This gas anchor included a 4-ft section of seal dividers placed on a latch sub. This is followed by a 2 3/8 in. x 4 ft internally plastic coated sub. Next, a downhole separator sub is installed on the 4 ft sub. Then, three joints of tubing are joined together with stainless steel bands.
The 2 3/8-in. tubing next to the 1-in. tubing was necessary because the assembly needed enough clearance in the 5 1/2 in., 23 lb/ft casing.
Because the equipment below the banded tubing would be submerged in stagnant fluid, the equipment was nickel plated prior to its installation.
TUBING
The existing 2 7/8 in., N-80 EUE tubing was used in the rod pumping design. The tubing ID was large enough to accommodate the 1.25-in. rods and the tubing provided sufficient burst resistance for the hydrostatic pressure of the fluid column.
The tubing also provided the strength to pull the tubing if the fiber glass rods parted and could not be fished. This is called a stripping job. In this job, the tubing is pulled out of the well until the top of the rod fish is found. At this point, the rods can be pulled to unseat the pump, and afterwards the rods and pump can be pulled from the tubing.
The highest load on the tubing during a stripping job is when the tubing is pulled out of the tubing-head slips. This force on the tubing string is calculated in the example box.
The calculation shows the worst case estimate of the load on the top joint of tubing if the rods had to be stripped out of the well. The joint yield strength for this tubing is 144,960 lb.
SHEAR TOOL
Fiber glass rods cannot be torqued over 100 ft-lb without damage to the fibers in the rod body. If the insert pump became stuck in the tubing or seating nipple, a means of releasing the rod string would be necessary.
An all steel rod string could be backed-off and pulled out of the well before the tubing was pulled to retrieve the pump. This releasing technique cannot be used with fiber glass rods.
Therefore, a shearing device is placed below the fiber glass rods so that the rods can be pulled from the tubing without using torque. The shearing value of the tool needs to be set so that the maximum short-term load on the top fiber glass rod, and the yield strength of the steel rods are not exceeded.
The shear tool must be large enough to handle the pump loading. A 26,000-lb shear tool was selected for the Barbara Fasken Fee C well.
PUMP
An 1.0625-in., heavy-wall insert pump was run in the well.
The insert pump can be changed without pulling the tubing string, and the heavy-wall pump provides sufficient strength for the anticipated hydrostatic pressure of the fluid column.
Because the pump joint was 2 3/8-in. tubing, only two heavy-wall insert pump barrels are available to fit inside the pump joint. These are an 1.062-in. and an 1.25-in. pump.
The 1.0625-in. pump was selected because the pump load would be less than for the 1.25-in. pump. With fiber glass rods, a lower pump load means more pump overtravel.
For selecting a 36-in pump barrel, the length calculation considered the plunger length, maximum stroke length of the pumping unit, and maximum pump plunger overtravel.
Maximum plunger overtravel will occur when pump load is lowest. This is at start-up when the casing fluid level is highest.
A mechanical hold down was selected to seat the pump.
This hold-down device is easier to release than a cup-type hold down.
Pump specifications and materials are listed in Table 1.
INSTALLATION
The installation procedure followed API's RP11BR.5 This publication includes two topics that were important to the success of this installation: corrosion control by chemical treatment, and sucker-rod joint makeup utilizing circumferential displacement.
The corrosion control process began by lubricating the sucker-rod pins with a corrosion inhibitor during makeup. Sucker-rod couplings and pin ends can become invaded with well fluids when the rods are run into the well.6
The hydrostatic pressure of the fluid column forces well liquids into the threaded portion of the rods. If the well fluids are corrosive, the fluids will degrade the unprotected portion of the rod string.
After the rod string was run, the well was batch treated with corrosion inhibitor. When rods and tubing are pulled from a well, the rods and tubing are subjected to atmospheric corrosion. Corrosion inhibitor treatments before or after a pulling job will reduce the effects of this type of corrosion.
Also, a corrosion maintenance program batch treats the well with 2 gal of corrosion inhibitor every 2 weeks. Coupon tests indicate metal losses of less than 0.5 mils/year (mpy). The maintenance program has not been altered since the installation.
For making up the sucker rods, API's guide on circumferential displacement was followed. Power tongs applied the torque required for each size of rod coupling. One should note that the rod couplings were made up slowly to avoid overtorqued rods. The tongs were rechecked for calibration every 25 rods.
Also, it should be noted that the displacement guide is different for new and used Grade D rods.
On site supervision of the installation procedure and close adherence to API's recommendations for handling sucker rods is considered key to the long-term success of this rod pumping installation.
A pump-off controller was also installed and calibrated to ensure that the well shuts down at pump-off. When using fiber glass rods, a pump plunger pounding fluid will result in an end fitting pull out or body failure.
Currently the well is pumping 55% of the time.
ECONOMICS
After the pumping system was changed out, the operating expenses on this lease decreased from $7,500/month to $1,350/month. Therefore, the $100,000 cost of installing the sucker-rod pumping system was paid out in 1.4 years.
The reduced operating expense also significantly reduced the economic limit of the property from 450 bbl of oil/month to 64 bbl of oil/month. This lower economic limit represents an additional 190,000 bbl of recoverable reserves (Fig. 6.
REFERENCES
- Lea, J.F., and Minissale, J.D., "Efficiency of Artificial Lift Systems," Thirty-ninth Annual Southwestern Petroleum Short Course, Lubbock, Tex. Apr. 23, 1992.
- Gibb, S.G., and Nolen, K.B., Nabla Corp.'s SROD predictive program.
- Gault, R.H., "Designing a Sucker Rod Pumping System for Maximum Efficiency," SPE Production Technology Symposium, Lubbock, Tex., November 1985.
- Clegg, J.D., "Another Look at Gas Anchors," Thirty-sixth Annual Southwestern Petroleum Short Course, Lubbock, Tex., Apr. 1920, 1989.
- API RP11BR, "Recommended Practices for Care and Handling of Sucker Rods," Eight edition.
- Winfield, M.S., "Sucker Rod Lubricants," Paper No. 435, Corrosion/90, Las Vegas, Nev., 1990.
Copyright 1993 Oil & Gas Journal. All Rights Reserved.
