Hu Yongquan, Wu Zhijun
Southwest Petroleum Institute
Nanchong, China
sucker rod pump can efficiently dewater gas wells if the separation coefficient is sufficiently high. It is not sufficient only to know if the system meets the criteria of rod string stress (RSS), horsehead load (HL), and crankshaft torque (CT).
Because gas wells have high gas/liquid ratio (GLR), a pump can easily gas lock and prevent discharge of water from the well.
Water production
Reservoir energy and gas production rate gradually decrease as a gas field depletes. Often, water production begins after producing a certain amount gas.
In the Sichuan gas field, to try to increase gas production rates, a series of tests was done on pumping units for lifting water while producing gas. These units all met the design criteria of RSS, HL, and CT but in some cases were not able to discharge water from the well.
To determine the reason for this, pump fill factors were first calculated and analyzed. Although the pumps cannot discharge if gas-locked, efficiency can be improved by enhancing the separation coefficient.
General conditions
In one well, gas production began in July 1977 at a rate of 480,000 cu m/d (16.9 MMcfd). At the beginning, no water was produced. In February 1985, gas production decreased to 380,000 cu m/d, and water increased to a maximum of 56 cu m/d (352 b/d). In November 1986, the water production rate was generally 50 cu m/d, with a maximum of 86.6 cu m/d.
To increase the gas production rate, tests with pumping units began in 1990. The basic parameters (see nomenclature box)(26208 bytes) of a gas well in 1990 were:
- Well depth (H): 2,622.61 m (8,604 ft)
- Gas specific gravity (gg): 0.571
- Water specific gravity (gL): 1.0126
- Gas production rate (QG): 1,000-2,000 cu m/d (35-70 Mcfd)
- Water production rate (QL): 55 cu m/d (346 b/d)
- H2S in gas: 0.247 g/cu m
- Surface temperature (Ts): 293 K (20 F.)
- Temperature at pump setting (Ta): 348 K (347 F.)
- Static bottom hole pressure (Pws): 6.3 MPa (912 psi)
The design parameters of the pumping units were:
According to the design criteria, the pumping units met the RSS, HL, and CT criteria.
Problem analysis
Field operations confirmed that a pumping unit could meet the RSS, HL, and CT criteria but could not discharge water; thus, preventing the well from producing gas.
On Nov. 22, 1990, the production rate of water was 49 cu m/d (308 b/d), and pump efficiency was 55.8%. On Nov. 26, the well discharged water and produced gas simultaneously, but later the pump could not discharge water from the well. The casing pressure increased to 4 MPa (580 psi), and the sucker rod pumping units continued to run for 1 hr without discharging any water from the well.
On Mar. 7, 1991, the unit also would not pump water. To try to discharge water, the rod spacing was decreased from 0.55 to 0.30 m, but the pumping units still would not discharge water.
The fill factor (b) of a pump is defined as the ratio of inlet liquid volume of the pump in the up-stroke to the pump displacement volume (Fig. 1 (51930 bytes)). It is obvious that b reflects pump volume filled by liquid in the up-stroke. Higher b indicates more liquid volume is in the pump. If b is zero then the pump is definitely gas locked.
Fill factor
Based on a gas well producing 39,920 cu m of gas and 93 cu m of water in 91.83 hr before it ceased production, the following steps indicate the method for determining the fill factor:
1.VLs = 93 cu m and Vgs = 39,920 cu m
2.Pressure at the pumps standing valve (SV) cannot be obtained from Reference 1, which shows several equations for determining pressure at the standing valve. Because gas has a significant effect on the pressure, the equations are not suitable for a gas well. But because the flowing pressure loss is low in a high gas/liquid ratio well, about 429 cu m/cu m in this well, one can assume that the possible maximum pressure (PMP) at the pumps standing valve is equal to the static bottom hole pressure, 6.3 MPa in this example.
3.Based on the gas specific gravity, pseudocritical pressure (Ppc), and pseudocritical temperature (Tpc), the gas deviation factor at the pump is as follows:2
Ppc = 709.6 58.7 x gg = 676.1 psia
Tpc = 170.5 + 307.34 x gg = 346 R
Therefore, pseudoreduced pressure (Ppr) and pseudoreduced temperature (Tpr) are:
Ppr = PMP / Ppc = 1.35
Tpr = Ta / Tpc = 1.81
According to the Standing and Katz correlation, gas deviation factor (Z) is:
Z = 0.92
4.If the formation water volume factor (Bw) is assumed to be 1, the gas/liquid ratio (GLR) at the pump is calculated as follows:
Psc x Vgs/Ts = PMP x Vga/(Z x Ta)
Vga = Psc x Z x Ta x Vga/(PMP x Ts) = 692.7 cu m
VLa = VLs/Bw = 93 cu m
GLR = Vga/VLa = 7.448 cu m/cu m
5.The fill factor,3 b, for 0.55 m and 0.30 m rod spacing is as follows:
b = VL/Vp (1)
From Fig. 1:
Vp + Vs = Vg + VL (2)
therefore:
Vg = GLR x VL (3)
thus:
Vp + Vs = GLR x VL + VL (4)
And from Fig. 1:
VL = VL 2 Vs (5)
Substituting Equation 4 and 5 into Equation 1 gives:
b = (Vp + Vs)/(Vp x (1 + GLR)) Vs/Vp (6)
where:
Vp = 0.25p x D2 x (S e)
Vs = Traveling valve volume + standing valve volume + pump volume of rod spacing
In this pump, the traveling valve and standing valve volumes are far smaller than the pump volume of the rod spacing, therefore, these volumes can be neglected in calculating the total unused space as follows:
Vs = Pump volume of plunger space = 0.25p x D2 x Sr
If Sr = 0.55 m than from Equation 6:
b = 0.006
If the separation coefficients of the separator are 20%, 40%, 60%, 80%, and 90%, the homologous fill factors are 2.3%, 6.8%, 14.4%, 31.6%, and 51.3%, respectively.
For an Sr = 0.30 m, the similarly results are:
b =0.05
And with separation coefficients of separator of 20%, 40%, 60%, 80%, and 90%, the homologous fill factors are 7.8%, 12.0%, 19.3%, 35.4%, and 54.0%, respectively.
Analysis
The previous calculations show the basic way to improve fill factors for the pump to enhance the separation coefficient to a value at which the pumping unit can discharge liquid.
When the rod spacing is 0.55 m, surface GLR is as high as 429.2 cu m/cu m based on the previous producing data. Actually, pressure at the pump is certainly less than Pws, thus, GLR at the pump is greater than 7.448 cu m/cu m.
If the separation coefficient is less than 40%, the pump is gas-locked (b = 6.8%) and cannot discharge any water. In fact, on Mar. 8, 1991, the result of inspecting the pump confirmed that the downhole gas separation was blocked by iron oxide powder, so that the pumping unit could not discharge water.
If the separation coefficient in this well is 60%, the fill factor of the pump is as low as 14.4%. If the separation coefficient is 80%, the fill factor is increased to 31.6%.
If one is sure that the plunger will not collide with the standing valve, a reduction in the rod spacing will improve the fill factor, but this has little effect in gas wells with a high GLR.
For example, when the separation coefficient is 80%, a reduction of the rod spacing from 0.55 m to 0.30 m changes the fill factor only from 31.6% to 35.6%, but increases the possibility of the plunger colliding with the standing valve.
Reference
- Brown, K.E., The Technology of Artificial Lift Methods, Vol. 4, PennWell, 1984.
- Ikoku, C.U., Natural Gas Production Engineering, John Wiley & Sones Inc., 1984.
- Wang Hongxun, Zhang Qi, The Principles of Oil Well Production Technology (revised version), Petroleum Industry Publishing House, 1989.
The Authors
Hu Yongquan is a lecturer at the Southwest Petroleum Institute, Nanchong,
China. He is involved in conducting research on production, stimulation, and completions. Hu graduated from Southwest Petroleum Institute (SWPI) in 1985 with a BS, and he received an MS of petroleum engineering in 1988.
Wu Zhijun is a lecturer at the center for well completion techniques of the China National Petroleum Corp. He conducts research on drilling, completions, and production operations. Wu received BS and MS degrees in petroleum engineering from SWPI.
Copyright 1995 Oil & Gas Journal. All Rights Reserved.