Fall 2001 ASME/API Gas Lift Workshop,
November 12-13, 2001, Aberdeen.
PRESSURE PULSE ANALYSIS OF GAS LIFT WELLS
J.S. Gudmundsson, I. Durgut
Norwegian University of Science and Technology
J. Rønnevig, K. Korsan, H.K. Celius
Pressure pulse technology makes possible gas-liquid rate metering and flow condition analysis of wells and flowlines. The technology is used to make on-demand measurements, when production rate and wellbore flow condition data are needed. Pressure pulse testing over time (at intervals of hours, days, weeks) makes possible long-term monitoring of a range of well production parameters useful in reservoir, production and facilities management. Because pressure pulse technology is simple, flexible and gives highly-repeatable results, it can be used in smart wells and smart fields operations. Pressure pulse testing of gas lift wells can be used to locate operating valve and to analyze wellbore condition, thus eliminating the need for wireline pressure surveys. A simulation study of a typical gas lift well with valves at 1100 m, 1750 m and 2100 m depths was carried out. The use of pressure pulse technology to optimize gas lift operations was discussed.
Pressure pulse technology is for on-demand measurements of production rate and wellbore condition analysis. With the advent of high-quality pressure transducers and computer-based data acquisition systems, it has become possible to make practical measurements of rapid pressure transients. The widespread use of quick-acting valves in the oil industry to open, close and control pipeline and wellbore flows, has made it possible to harness the information contained in rapid pressure transients when valves are activated. Pressure is one of the easiest parameters to measure in the production of oil and gas. It can be measured in pipelines, flowlines and wellbores; at wellheads, chokes, manifolds and separators.
Pressure pulse technology uses the information contained in the pressure signal measured up-stream of a quick-acting valve in a multiphase flow channel, including onshore and offshore oil wells (Gudmundsson and Celius, 1999). Field testing on the Oseberg, Gullfaks and Snorre production platforms, has demonstrated the feasibility of pressure pulse technology for production testing and wellbore logging and monitoring.
In pressure pulse technology the pressure profile in a pipeline can be used to detect and monitor solid deposits. The pressure profile is obtained from pressure measurements at one location, immediately up-stream of a quick-acting valve. When the valve is activated, the up-stream pressure is measured, resulting in a pressure-time log. The pressure-time log is then converted into a pressure-distance log. The pressure-distance log gives the location and extent of deposits in a pipeline (Gudmundsson et al. 2001).
Pressure pulse technology can be used in gas lift wells for flow rate metering and flow condition analysis. For example, pressure pulse testing can be used to detect the open/closed condition of gas lift valves. A case study for such a gas lift well will be presented. Pressure pulse measurements are made at the wellhead; there is no need to enter the well (non-interventive technology). A pressure-time log is used to detect and monitor the wellbore flow condition from wellhead to bottomhole.
Pressure Pulse Generation
When a quick-acting valve in a multiphase pipeline is activated, a pressure pulse will be generated. The pressure pulse will propagate both up-stream and down-stream of the quick-acting valve. The magnitude of the pressure pulse will be governed by the water-hammer equation, also called the Joukowsky equation:
where ρ (kg/m3) represents the fluid density, u (m/s) the fluid flowing velocity and a (m/s) the
speed of sound in the fluid. The speed of sound in the fluid is equivalent to the propagation speed of the pressure pulse generated.
A typical pressure pulse technology set-up is shown in Figure 1a. It shows a quick-acting valve and two pressure transducers, A and B, up-stream of the valve. The pressure transients generated at locations A and B are shown in Figure 1b. The quick-acting valve generates a rapid increase in the pipeline pressure at locations A and B. The initial rapid pressure increase is the water-hammer, as given by the Joukowsky equation. The pressure pulse will arrive at location A before it arrives at location B. The time difference is the time-of-flight ！t (s) which can be used to determine the speed of sound in the flowing gas-liquid mixture:
The magnitude of the pressure pulse generated by a quick-acting valve can be measured immediately up-stream by a pressure transducer. In flow systems where the up-stream and down-stream pipes are sufficiently long, the pressure increase immediately up-stream of the quick-acting valve, will be the same as given by the water-hammer equation.
A pressure pulse travelling up-stream in a pipeline, will arrest (stop) the flow. The pressure pulse will travel up-stream the pipeline at the in-situ speed of sound. Therefore, the oil and gas will be brought to rest as quickly as the pressure pulse travels down the pipeline. In principle, when the pressure pulse has reached the end of the pipeline, the fluid velocity will be reduced to practically zero.
Mass Flow Rate
2The mass flow rate in a pipe of constant cross-sectional area A (m) can be obtained directly
from the Joukowsky water-hammer equation, when the sound speed is also measured. The 2mass flux G (kg/s.m) can be expressed as
3where ， (kg/m) is the gas-liquid mixture density and u (m/s) the mixture average velocity.
The mass flow rate m (kg/s) is given by the expression
Therefore, provided the sound speed a (m/s) and the water-hammer pressure increase are
known from measurements, the mass flow rate can be found directly from the relationship
The continuity principle dictates that the mass flow rate at the quick-acting valve is the same as the mass flow rate at other locations, including: downhole, separator and stock-tank. It means that the mass flow rate m (kg/s) measured in the pressure pulse method is exactly the same as the mass flow rate entering a wellbore in the pay-zone.
Flow rates in the petroleum industry are traditionally expressed in volumetric flow rates. The formation volume factors of oil, gas and water are used to determine the volumetric flow rates at various pressures and temperatures.
As the flow is brought to rest, the pressure loss due to wall friction will be made available. That is, the pressure drop due to gas-liquid mixture flow in the pipeline, will be released. This frictional pressure drop will propagate continuously to the quick-acting valve and can be measured and is often called line-packing. The gradual pressure increase (pressure gradient) after the initial water-hammer in Figure 1b is the line-packing.
Frictional pressure drop in pipes is governed by the Darcy-Weisbach equation:
where f (dimensionless) is the friction factor, ΔL (m) pipe length, d (m) pipe diameter, ρ 3(kg/m) fluid density and u (m/s) fluid velocity. The ΔL is not the same distance as the ΔL AB
used to determine the speed of sound. The Darcy-Weisbach equation as shown here holds for single-phase laminar and turbulent flow. In principle, the equation can be extended to hold also for multiphase flow.
The density of a gas-liquid mixture is given by the relationship:
where α (dimensionless) is the void fraction and the subscripts stand for M (mixture), G (gas)
and L (liquid). In hydrocarbon production the liquid-phase will often consist of oil and water.
The speed of sound in homogeneous gas-liquid mixtures a is given by the traditional Wood M
equation, expressed as:
and a and a the speed of sound in gas and liquid, respectively. GL
An in-house computer program was used to estimate the speed of sound in the gas-oil mixture flowing at 60 bar and 90 bar in an offshore well. The wellhead pressure was about 90 bar and the separator pressure about 60 bar. For two-phase mixtures the computer program gives the same result as Wood’s equation.
The results of the speed of sound calculations are shown in Figure 2 with void fraction from 0 to 1. The speed of sound in pure liquid is high and decreases dramatically with small amounts of gas. In the void fraction range 0.2-0.8 the sound speed remains relatively constant. As the void fraction increases from 0.8 to pure gas, the sound speed increases. Model calculations show that as the pressure increases, the sound speed in gas-liquid mixtures also increases.
Typical pressure pulse values in pipelines and oil producing wells will be in the range 2-10 bar, depending on fluid composition and flow rates (Gudmundsson et al. 2001). Pressure pulse rate metering was qualified during long-term testing on the Gullfaks B platform in 1999. A standard 6" hydraulically operated ball valve with a modified actuator was used to close the wells. The valve was designed for fail-safe-close operation; that is, kept open by the use of hydraulic pressure (when fluid bled off the pressure fell and a spring closed the valve). The valve closing time was determined by the combined effect of the spring force and the drainage capacity. The valve closing time on Gullfaks B was around 0.2 seconds during pressure pulse measurements.
The Gullfaks B qualification test lasted for 10 months, and a total of 800 closing operations were performed using the same quick-acting valve. The reservoir is poorly consolidated, and the wells produce some sand. No leakage was observed across the valve during the testing period. Based on the Gullfaks B experience, and discussions with valve manufacturers, a valve maintenance period of at least 5 years is achievable when pressure pulse is installed on individual wells.
Wells and Pipelines
Multiphase flow in land-based and offshore wells and pipelines depends on many factors, including the pressure, volume and temperature behaviour of the fluid mixtures involved. It is convenient to illustrate the use of pressure pulse technology in multiphase flow by assuming several of the main factors as constant. Later, in practical situations, such assumption can be relaxed and the various effects can be taken into consideration. For illustration purposes the following assumptions were made:
~ Single-phase flow.
~ Constant diameter.
~ Constant friction factor.
~ Constant flowrate.
~ Constant in-situ speed of sound.
~ Constant fluid viscosity.
Given the above assumptions, the line-packing measured at a quick-acting valve after full/complete closing, will increase with time. Furthermore, assuming that the quick-acting valve closes instantaneously, the pressure increase with time for such conditions is illustrated in Figure 3a. For any point A (not the same A as illustrated in Figure 1) the pressure measured represents the well or pipeline line-packing the distance ΔL up-stream:
where Δt (s) is the time. The factor 0.5 is applied because the pressure pulse must first travel the distance A and then back to the quick-acting valve.
The assumption of a constant well or pipeline diameter can be relaxed to illustrate the situation where the diameter has been reduced in a certain interval. The diameter reduction is abrupt and significant and exists for some distance, until the diameter expands abruptly and significantly. The pressure increase with time for such a condition is illustrated in Figure 3b. The point C represents the distance from the quick-acting valve to the reduction in well or pipeline diameter, and the point D represents the distance from the valve to the resumption of full diameter. Such a reduction in diameter the distance CD may result from the deposition of solids in the particular interval.
The assumption of a constant friction factor can be relaxed to illustrate the situation where the friction factor increases in a certain interval. An increase in friction factor will result in similar effects as a decrease in diameter, as evident from the Darcy-Weisbach equation. The increase in friction factor increases the frictional pressure gradient in the interval, as illustrated in Figure 3c. The point E represents the distance from the quick-acting valve where friction increases, and the point F represents the distance from the valve where friction decreases. It needs to be recognised that the deposition of solids in a certain interval and resulting in a reduced diameter, may also be accompanied by a change in friction factor.
Figures 3a-3c illustrate the increase in water-hammer pressure when a quick-acting valve is closed, and the subsequent gradual increase in line-packing pressure with time. The figures illustrate simplified situations, and the points A, C, D, E and F represent for each situation a particular distance ΔL. To calculate this particular distance, fluid flow equations and fluid properties need to be known. In single-phase flow of fluids with constant pressure-volume-temperature (PVT) properties, the calculations are simple and explicit. In multiphase flow of fluids with variable PVT-properties, however, the calculations are more involved and implicit.
The following steps describes how the distance ΔL might be calculated for the particular
situation illustrated in Figure 3b where the point C represents the distance to the start of solids deposition (reduction in diameter, for example) in a well or pipeline:
1. A pressure pulse test is made and the mass flowrate of the gas-liquid mixture flowing
at the quick-acting valve is calculated from the water-hammer equation, and the
temperature is measured.
2. The pressure-volume-temperature properties of the gas-liquid mixture flowing at the
quick-acting valve are assumed known from standard oilfield practices, based on
measurements and/or established correlations.
3. A commercial or in-house pipeline flow simulator is then used to calculate the well or
pipeline pressure and temperature from the quick-acting valve and up-stream,
including fluid densities and void fraction.
4. The speed of sound in the flowing gas-liquid mixture is then calculated piecewise
from the quick-acting valve and up-stream, using fundamental relationships and the
flow simulation results.
5. The time-scale in Figure 3b is converted to distance in a piecewise manner using the
Δt. relationship ΔL = 0.5 a
The above calculations can be carried out using data and models that range from simple to comprehensive. The accuracy of the calculations can be improved by additional measurements. For example, pressure measurements from somewhere up-stream can be matched to the arrival of the pressure pulse. And the known locations of changes in diameter and other features can be matched to their appearance in the line-packing signal measured at a quick-acting valve. Similarly, up-stream temperature measurements can be used to improve the accuracy of pressure profiles in wells and pipelines; either point measurements or distributed measurements.
On-demand pressure pulse tests and measurements have been made in multiphase production wells on several platforms in the North Sea. The tests have shown that the theories expressed by the Joukowsky equation (water-hammer), the Darcy-Weisbach equation (line-packing) and the Wood equation (pressure wave propagation), are applicable in practical oil and gas production situations. Pressure pulse technology has been demonstrated to be cost-effective, flexible and highly-repeatable.
A recent offshore test illustrates the flexible nature of pressure pulse technology. The technology can be applied in situations ranging from a limited-installation to an extensive-installation. A limited-installation test was carried out in 2001 on a multiphase flow oil producer in the North Sea, called Well A. The wing valve was used to generate the pressure pulse and the pressure transient was measured with the existing wellhead transducer. The data were sampled at 1 second intervals. Normally, two fast sampling pressure transducers spaced 3-5 m are used to determine the speed-of- sound at line conditions. In the case of Well A, the speed-of-sound was determined from the pressure reflections from a change in tubing ID (inside diameter) from 4.89 inch to 3.55 inch at 650 m depth.
The measured wellhead pressure is plotted with time in Figure 4 (pressure-time log). The hydraulic wing valve was used to stop the flow. The valve starts to close at approximately 7 seconds, and is fully closed at approximately 10 seconds. Quick-acting hydraulically activated valves can be closed in about two-tenth of a second. Most valves in petroleum production operations can be closed in a couple of seconds; importantly, most of the closing action occurs after about 80% of the stem movement.
The size of the pressure pulse is 0.4 bar in Figure 4. The reflection from the tubing diameter transition point starts emerging at the wellhead transducer at about 15 seconds and is fully developed at approximately 20 seconds. This gives an average acoustic velocity of about 140 m/s in the upper tubing. There is more noise in the signal in the 4.89 inch tubing than in the 3.55 inch tubing. However, it is clearly seen that the line packing gradient is greater in the smaller diameter tubing, because of the higher frictional pressure drop. The following data were reported for Well A:
~ Measured depth 4440 m
~ Vertical depth 2270 m
~ Produced oil 35 API gravity 33~ Gas oil ratio (GOR) 24.8 Sm/m
~ Gas gravity 0.89
~ Wellhead pressure 67 bara 3~ Liquid production rate 1167 Sm/day
~ Water cut 66.8% 33~ Produced GOR 635 Sm/m.
~ Wellhead temperature 76 ：C
~ Bottom hole pressure 233.7 bara
~ Bottom hole temperature 100 ：C
~ Void fraction at wellhead 0.63
The well was choked before the pressure pulse test was made; that is, the flow rate was reduced below the normal production rate. An in-house transient pressure simulator was used in combination with a commercial steady-state wellbore flow model and an in-house acoustic velocity model to calculate the pressure with time. The simulation result is shown as a solid line in Figure 4. It was found that choking of the well prior to the measurement reduced the flow rate to about 20% of the test separator rate. Without the choking, the size of the pulse would have been about 2 bar.
The limited-installation setup used on Well A has also been applied to other wells with good results. The measurement accuracy will be less than can be achieved with two fast sampling transmitters and quicker valve closing. All the same, the flow rate determination has been within 15% of the test separator reading, and the mixture density was determined with even better accuracy. This illustrates the flexibility of the pressure pulse technology, and shows that the method in many situations can be used with only moderate modifications of existing design.
Gas Lift Analysis
Computer codes that simulate the propagation of pressure pulses in wellbores can be used to illustrate the use of pressure pulse technology to detect injecting gas lift valves. Gas injection changes the fluid and flow properties in the tubing, and in turn, the propagation and reflection characteristics of pressure pulses. This is illustrated through simulations on Well B (an example gas lift well). Input data to the simulations are given in the following:
~ Measured depth 3500 m
~ Vertical well
~ Tubing inside diameter 0.1005 m
~ Produced oil 32 API gravity
~ Gas gravity 0.85
~ Water gravity 1.103 33~ Gas oil ratio (GOR) 50 Sm/m
~ Wellhead pressure 50 bara 3~ Liquid production rate 400 Sm/d
~ Water cut 50 % 3~ Gas injection rate 100 MSm/d
~ Valve locations: 1100 m, 1750 m and 2100 m
Well B used in the calculations in shown in Figure 5. The well is fitted with three gas lift valves, located at 1100 m, 1750 m, and 2100 m respectively. The production enters the tubing at the bottom of the well, which is at 3500 m. The quick-acting valve is located at the wellhead, and was assumed to take 0.5 second to close. The pressure propagation is recorded at a fast sampling transmitter located immediately upstream of the quick-acting valve.
The in-house pressure transient simulator of pressure pulse propagation needs steady-state profiles of mixture density and acoustic velocity along the wellbore. A commercial steady-state wellbore flow simulator was used to calculate the density profile, and an in-house acoustic velocity model provides the acoustic velocity profile. The mixture density, void fraction and acoustic velocity in Well B are shown in Figure 6 for gas injection through the valve at 1100 m depth. Even a small amount of gas shows up as a change is the acoustic velocity. Below the bubble point, which is located at approximately 2800 m, the acoustic velocity increases only slightly with depth.
The pressure pulse and line packing simulation with the 1100 m valve injecting gas are shown in Figure 7a. The gas injection rate is 100 MSm3/d (1.16 kg/s), and the liquid production rate is 400 Sm3/d (4.58 kg/s). The initial pressure increase from 50 bar to about 51.3 bar is the pressure pulse and the more gradual pressure increase after that is the line-packing pressure. The change in pressure appearing at approximately 17 seconds on the line packing curve is the response from the gas lift valve. This is more clearly seen on the pressure derivative plot, which is shown together with the pressure response curve in Figure 7a. The drop in pressure at about 25 seconds is the pressure response from the bottom of the well.
The results for all three cases simulated are shown in Figure 7 (a, b and c). The production rate and the gas injection rate is the same in all simulations. Deeper valve location is recognized as a time-shift in reflection on the pressure derivative plot. At higher gas injection rates the pressure pulse will dampen more. Field testing is needed to demonstrate the use of pressure pulse technology in gas lift operations.
A muti-pointing situation has also been simulated. Input data was the same as in the preceding simulations except for the injection rate, which was split fifty-fifty between the valves at locations 1750 m and 2100 m. The result is shown in Figure 8. Multi-pointing is recognized by several reflections on the pressure derivative plot.
Analysis of the line-packing pressures shown in Figure 7 and 8 make it possible to assess the status of gas lift valves, and to identify which valves are injecting gas. Such analysis will include the measurement of flowrate by pressure pulse testing (Gudmundsson and Celius 1999).
The use of pressure pulse technology makes possible on-demand measurements of gas-liquid flow rate and flow condition analysis of the wellbore, from wellhead to bottomhole. The testing of pressure pulse technology offshore has focused on gas-liquid flow rate measurements and increasingly on wellbore logging (condition analysis).
The flexibility of pressure pulse technology has recently been demonstrated on an offshore platform (Well A) using already installed wing valve, one wellhead pressure transducer and data acquisition system. Limited-installation systems can be used to measure and monitor parameters of special interest, for example gas and/or water break-through.
In-house tools to model rapid pressure transients in wells and to calculate the speed of sound in gas-liquid mixtures, and a commercial wellbore flow simulator, were used to study a typical gas lift well (Well B). Pressure pulse analysis can be used to identify which gas lift valve is active. In the operation of gas lift wells pressure surveys are commonly needed to identify the open-closed condition of gas lift valves. A simple pressure pulse test has the potential to give the information needed, based on wellhead measurement only. Field testing of pressure pulse technology on gas lift wells is needed.
In addition to condition analysis of gas lift valves, pressure pulse technology can be used to optimize gas lift production systems. For example, in situations where several wells are gas lifted and the availability of compression power is limited. To achieve optimum oil production, the available injection gas needs to be distributed to the wells giving most oil. Pressure pulse technology can be used to measure the production rate of each well in quick succession, to make possible necessary adjustments in gas injection rates.
Gudmundsson, J.S. and Celius, H.K. (1999): Gas-Liquid Metering Using Pressure-Pulse Technology, SPE 56584, Annual Technical Conference and Exhibiton, Houston, 3-6 October.
Gudmundsson, J.S., Durgut, I., Celius, H.K. and Korsan, K. (2001): Detection and Monitoring thof Deposits in Multiphase Flow Pipelines Using Pressure Pulse Technology, 12 International
Oil Field Chemistry Symposium, April 1-4, Geilo.
Figure 1a – Pressure pulse set-up for a pipeline, showing quick-acting valve and pressure
Figure 1b – Pressure pulse at locations A and B up-stream a quick-acting valve.