Paul Branagan
Branagan & Associates
Las Vegas, Nev.
Bruce L. Knight
Marathon Oil Co.
Littleton, Colo.
John Aslakson
Gas Research Institute
Chicago
Michael L. Middlebrook
CER Corp.
Las Vegas, Nev.
A study has concluded that production logging tools employed to evaluate multiphase horizontal well production behavior should be carefully screened as to their response characteristics in fully segregated, two phase flow.
The study, performed at Marathon Oil Co.'s petroleum technology center in Littleton, Colo., indicated that gas in highly deviated well bores segregates rapidly in the presence of water, creating a downhole environment that produces sporadic responses from full bore and diverter spinners as well as density and holdup tools.1
Gas Research Institute (GRI), as part of its horizontal gas well completion technology program, initiated the full scale laboratory study to determine the severity and consequences of multiphase flow on tool response from horizontal well production.
BACKGROUND
As the petroleum industry accelerates the use of horizontal wells, completions become more complex, and acquisition of accurate production log data becomes critical in assessing reservoirs and production.
Furthermore, reservoir and stimulation engineers and log analysts are confronted with new and demanding analysis and design concerns as horizontal completions are augmented through hydraulic fracturing. To properly determine where to initiate stimulation treatments, engineers and analysts must determine well bore fluid entry points and specify the type and production rates of individual fluids.
Historically, production logging has been the most common technique to determine liquid and gas entry points, assess fundamental production mechanisms, and evaluate complex well bore fluid distributions.
Production log analysis relies on the integration of data from spinners and holdup tools, plus sensors that measure density, pressure, and temperature. These tools and data are needed to develop a logical depiction of downhole flowing conditions and relative production magnitudes. Most production logging tools are designed to respond to single phase or highly mixed relatively homogeneous fluid environments.
In vertical wells, measurement of flow conditions is straightforward. Fluid flow is best described by a uniform distribution of gas bubbles slipping upward through the liquid phase. However, in highly deviated and/or horizontal wells, rapid phase segregation creates a situation in which logging tool response becomes sporadic or uncertain.
TEST OBJECTIVES
GRI's full scale laboratory study investigated the behavior of two phase (gas/ water) flow in deviated wells and the corresponding performance of various production logging tools in that complex environment. The objectives were to:
- Compare the performance of conventional production logging tools in vertical and deviated well bores.
- Assess tool response to a variety of gas water flowing conditions.
- Provide independent visual observations of the complex multiphase pipe flow regimes.
The strategy employed an existing large scale flow loop facility that permitted the selection of various well bore deviation angles, and control gas and water flow rates. A video recording system was used to observe the actual well bore flow patterns surrounding the tool being tested.
TEST FACILITY
The field scale flow loop facility simulated downhole flowing conditions for a variety of gas/water mixtures and flow rates.
A 40 ft long, 4.5 in. ID tiltable, clear acrylic tube (Fig. 1), was used to simulate the well bore. The equipment, shown in a condensed schematic in Fig. 2, is at Marathon's petroleum technology center.
The apparatus permits a test tool (in this case a production logging tool string) to be placed in the flow loop. Various combination of liquids and gases then can be injected to simulate well bore flowing conditions.
Injection rates are computer controlled and the resulting tool responses are monitored, analyzed, and stored in a PC based computer system. To supplement production tool sensor data, a picture of the area surrounding the tool string provides a visual indication of the complex flow patterns that surround the tool.
The production logging tool string examined in this study (Fig. 2) consisted of:
- Full well bore spinner
- Diverter spinner
- Densitometer
- Holdup (capacitance) tool.
EXPERIMENTAL PROCEDURE
Other authors have characterized the two phase flow regimes that exist in vertical,2-3 and deviated4 6 well bores. Typically, the flow patterns are classified into the following four major categories:7
- Stratified (smooth and wavy)
- Intermittent (slug and elongated bubbles)
- Annular
- Dispersed bubble.
Fig. 3 shows the basic flow patterns for horizontal pipes as well as a model-derived graph that delineates flow patterns for various superficial gas and water velocities. With these patterns and the pumping constraints of the test facility as a basis, gas and water injection rates that typify actual field conditions were included in a test matrix totaling about 200 different combinations.
The range of test velocities (Fig. 3) encompasses stratified and intermittent flow patterns.
The maximum gas injection rates for these tests was 12 cfm (17.2 Mcfd at atmospheric pressure or 1.0 MMscfd for a bottom hole flowing pressure of 1,000 psi). This rate would in practice maintain the vertical portion of the well bore unimpeded by liquids because 1.0 MMcfd should be more than adequate to continually lift and unload any liquid buildup that might occur in vertical tubing. Thus, the range of gas flow rates selected for the test matrix was 0.1 Mcfd to 1.0 MMcfd and was typical of actual field conditions.
Tests were designed to include variable injection rates as described above for four different inclination angles (0, 10, 45, 85), and three general flowing categories (single phase, gas only in water, and two phase).
SINGLE PHASE FLOW
Single phase flow rate cases were used as a base comparison for the more complex two phase cases. Fig. 4 is a composite diagram showing the diverter spinner response to water injection rates ranging from 50 to 750 bw/d for well bore inclinations of 0 and 85. Similar response data are shown in Fig. 5 for the full well bore spinner. Note that for both spinner types, the responses to liquid flow were essentially linear, except for some small disturbances for the full well bore spinner at the low end of the water rate curve in the 85 case.
Diverter spinner responses for the single phase gas flow rate cases appeared to be reliable for inclinations of 0 and 85 (Fig. 6). The full well bore spinner response was too small to measure for these gas flow rates.
Both the densitometer and holdup tools provided proper values for both of the single phase cases at 0 and 85.
TWO PHASE FLOW
The second portion of the study evaluated tool response when both phases, gas and water, were present in the well bore. In the initial cases, gas was flowed at varying rates through a well bore filled with water. This simulates field conditions where the well is not necessarily producing water but the well bore nonetheless has residual water lying on the bottom. Fig. 7 compares the diverter spinner response for a vertical well with gas flowing through water and the single phase gas case described earlier.
In both situations, the response of the diverter spinner is well behaved and linear with an offset increase for the water-filled case. This offset is probably due to the elevation in the superficial gas velocity because of the buoyancy of the gas in the water.
Diverter response to gas flowing thorough a column of water was well behaved and linear for vertical and well bore inclinations of 10 and 45.
However, for the near horizontal (85) case with gas flowing through a water column, the response of the diverter spinner exhibited a significant threshold before becoming linear. These diverter response data for gas flowing through a water column are shown in composite form in Fig. 8.
The video explicitly revealed the well bore flow regimes that existed during these tests, particularly in the two phase cases where gas and water were clearly segregated. Well bore fluids in all but the 85 deviation case were seen to be well mixed with numerous small gas bubbles slipping through the liquid phase that was punctuated by intermittent slug flow.
The diverter spinner response appeared to be well behaved in these flow regimes, even in the presence of slug flow. However, in the 85 case, the diverter response was marked by a significant threshold that (as seen in the video) was due to significant gas leaking around the diverter shroud. In the 85 case, complete gravity segregation was seen to occur rapidly (i.e., within several feet of the injection ports) as gas migrated to the top side of the pipe and slugged its way up the well bore.
The video showed that the slugging process, which occurred to some degree for all angular deviations, created rather large eddy currents and liquid fallback around the full bore spinner, causing sporadic blade rotation reversals and thus abnormal tool response. Fig. 9 shows the full bore spinner response data and a rendition, based on the video, of the flow patterns that created the abnormal tool response.
Densitometer response in the 85 case was found to be inconsistent with the vertical response for the two-phase cases and appears to be a result of phase separated gas slugging. As gas slugs were seen to pass across the densitometer sensing elements, the density reading alternated between liquid and gas densities. Similar erratic responses were observed from the holdup tool in the presence of gas slugging for the highly deviated cases.
The final set of tests included flowing both gas and water through the flow loop. Results of the diverter flowmeter for the 85 deviation case with water flow rates of 50 and 750 bw/d are shown in Fig. 10. Although these data reveal some improvement in reducing the gas only threshold (Fig. 8), with increased water flow rates, the response was still nonlinear. Gas slugging and bypassing the diverter shroud appear to be at the root of this abnormal behavior.
RECOMMENDATIONS
The results of these full scale laboratory tests indicate intrinsic weaknesses in the application of certain production logging tools to two phase flow analysis in deviated well bores. Both the full bore spinner and the diverter spinner responses were adversely affected by rapid gas phase segregation that created slugging.
This abnormal behavior was exaggerated in the highly deviated well bore cases and confirms similar performance problems from recent field cases. Similar response difficulties were encountered for the density and holdup tools, especially when gas was seen in the video to be slugging by the active area of the tool.
Visual observations of the complex two phase flowing conditions were captured on a video tape and identified specific flow patterns that caused erratic tool response.
Diverter spinner response was found to be linear, from vertical to well bore deviations of 45. At 85, however, slugging and gas leakage around the diverter shroud created nonlinear tool response, making interpretation of flow behavior uncertain.
The full bore spinner response was found to be erratic for all deviated well bore cases, principally because of eddy currents and liquid fallback at the spinner blades created by gas slugs passing the tool.
Density and holdup tool responses were found to be inconsistent for the vertical and deviation cases. This inconsistency was also found to be the result of gas slugging through the active sensing elements of the tools, creating, in effect, a single phase environment (either entirely water or entirely gas, depending on the size and duration of the gas slug).
Improvements in the sealing mechanism of the diverter shroud with the well bore would certainly tend to reduce gas leakage and improve diverter tool performance. However, the spinner blades are probably subjected to the same eddy currents and liquid fallback phenomena that were seen in the video and plagued the response of the full bore spinner. Therefore, further testing of diverter spinners that includes visual observation of the spinner blades would help determine the exact nature of spinner blade response to specific flow regimes in the small confines of the diverter channel.
Focused tools, such as density and holdup, that rely upon small samples of well bore fluid acquired at the center of the well bore will continue to be subjected to periods of essentially single phase fluid flow, either gas or water, during slugging in highly deviated well bores.
The resulting response will depend on the frequency and geometry of each gas slug. Dynamically altering the time constants and sensor profiles for focused tools or mixing the well bore fluids in the vicinity of these tools should tend to improve tool response. Mixing of the well bore fluids prior to passing through the spinner tools would also tend to improve their response.
It is clear from this research that complex two-phase (gas/water) flow in horizontal well bores creates erratic tool responses that can result in serious errors in the interpretation of production processes.
REFERENCES
- Knight, B.L., "An Experimental Comparison of Production Logging Tool Responses in Vertical and Deviated Wet Gas Wells," GRI Topical Report No. GRI-92/0386, September 1992.
- Anasari, A.M., Sylvester, N.D., Shoman, O., and Brill, J.P., "A Comprehensive Mechanistic Model for Upward Two Phase Flow in Well bores," SPE 20630, September 1990.
- Brill, J,P., and Beggs, H.D., Two Phase Flow in Pipes, Fifth Edition, 1988.
- Scott, S.L., and Kouba, G.E., "Advances in Slug Flow Characteristics for Horizontal and Slightly Inclined Pipelines," September 1990.
- Hill, T.J., and Wood, D.G., "A New Approach to the Prediction of Slug Frequency," September 1990.
- Xiao, J.J., Shoman, O., and Brill, J.P., "A Comprehensive Mechanistic Model for Two Phase Flow in Pipelines," September 1990.
- Brill, J.P., and Arirachakaran, A.J., "State of the Art in Multiphase Flow," JPT, May 1992.
- Turner, R.G., Hubbard, M.G., and Dukler, A.E., "Analysis and Prediction of Minimum Flow Rate for the Continuous Removal of Liquids from Gas Wells," JPT, November 1969.
- Robertshaw, S.E., and Peach, S.C., "Well illustrates challenges of horizontal production logging," OGJ, June 15, 1992.
Copyright 1994 Oil & Gas Journal. All Rights Reserved.
