You must log in to edit PetroWiki. Help with editing

Content of PetroWiki is intended for personal use only and to supplement, not replace, engineering judgment. SPE disclaims any and all liability for your use of such content. More information


Difference between revisions of "Tracer testing in geothermal reservoirs"

PetroWiki
Jump to navigation Jump to search
(No difference)

Revision as of 08:12, 17 September 2013

Tracers are used in geothermal reservoir engineering to determine the connectivity between injection and production wells. Reinjection of spent geothermal fluid is nearly universal to

  • Address environmental concerns
  • Provide reservoir pressure maintenance
  • Improve energy extraction efficiency

Because injected fluids are much cooler than in-situ fluids, knowledge of injectate flow paths helps mitigate premature thermal breakthrough. As in other applications of tracer testing, the goal of the tracer test is to estimate sweep efficiency of a given injection pattern.[1] Because geothermal systems tend to be open, tracer tests can also be used to estimate the extent of recharge/discharge or total pore volume.[2][3] Currently, however, the primary use of geothermal tracers is to estimate the degree of connectivity between injectors and producers. That information is subsequently used to develop an injection program that either minimizes or postpones injection returns in production wells while providing pressure maintenance.

Geothermal tracers

Because geothermal reservoirs are not usually developed on regular well spacing, well pairs may exhibit weak connectivity, and tracer tests must be conducted over long times, using large volumes of tracer to overcome thermal decay and dilution effects. For these and other reasons, extensive work has been invested in evaluating so-called "natural tracers." These can be thought of as compounds that are present in geothermal fluids naturally and whose concentrations may change during production and injection and may therefore be used to trace injectate. Examples of natural tracers include chloride,[4] ammonia,[5] and various stable isotopes of water.[6][7][8]

Artificial tracers have also been used extensively to determine flow paths in geothermal reservoirs. Tritium was the first artificial tracer used to trace geothermal injectate.[9] Since the early 1990s, various new compounds have been evaluated for use in geothermal reservoirs. Liquid-phase tracers have evolved from carboxylic and benzene sulfonic acids[10] to polyaromatic sulfonates,[11] which are stable thermally at temperatures greater than 300°C and have detection limits in the range of 102 parts per trillion (ppt). Vapor-phase tracers have evolved from chlorofluorocarbons used in the early 1990s to hydrofluorocarbons in the late 1990s.[12] To date, criteria for selection of tracers focus on:

  • Thermal stability
  • Low background concentrations
  • Low detectability
  • Being environmentally benign

More recently, issues such as sorptivity and volatility have been recognized as equally relevant characteristics that influence analysis.[13]

Tracer tests have been conducted for over 25 years in geothermal fields, including early work in:

  • New Zealand[14]
  • The Geysers in Northern California[9]
  • Lardarello in Italy[15]
  • Various Japanese fields[16]

In the last decade, more than 50 tracer tests have been conducted worldwide in geothermal fields.

Interpretation methods

Early workers in the field recognized that tracer tests could be used quantitatively to evaluate volumetric sweep efficiency of an injection program. Lovekin and Horne[17] applied optimization methods to maximize the residence time of injectate. This involved minimizing a tracer breakthrough function.

Vol6 page 0417 eq 001.png....................(1)


where the cij are referred to as the arc cost function for the travel arc between a given injection and production well pair (e.g., a streamline), and qriis the injection rate for injector i. The cost function is related to operational and geologic information for the field, including:

  • Tracer first arrival and peak arrival times
  • Horizontal distances and elevation differences between wells
  • Injection and production rates

The method was applied to optimizing production operations at the Wairakei Field in New Zealand.[17]

In 1991, Macario extended the previous work to use a natural tracer, chloride, to optimize reinjection in the Philippine field, Palinpinon. Shortly after commissioning the power plant in 1983, an increasing trend of chloride in the production wells was observed. This was interpreted as evidence of rapid return of reinjected fluids to the production sector of the field.[18] Because the chloride trend is associated with all injectate (i.e., not a specific injector), Macario[19] developed a linear combination method that expresses the chloride concentration as a linear combination of the injection wells active during the time interval considered. Produced chloride for a given well, Clp, is expressed as a linear function of the chloride injection rates.

Vol6 page 0418 eq 001.png....................(2)

The coefficients ai are coefficients of correlation between a given producer and injector. A large coefficient implies strong production contribution from a given injector. These coefficients can subsequently be used in the arc cost function. These methods appear to work well if there is operational flexibility to use the appropriate wells and work equally well for either natural or artificial tracers.

Noting that fluctuations in injection rates manifest themselves as changes in produced chloride concentrations over and above the underlying trend in time, Sullera and Horne[4] applied wavelet analysis to two geothermal fields:

  • Palinpinon in the Phillipines
  • Dixie Valley in Nevada

The chloride production data and injection rates are decomposed into progressively lower-frequency detail, and multiple regression techniques are applied to identify the degree of connectivity between individual injectors and producers. Care must be taken to avoid decomposing the signal too far; however, Sullera and Horne[4] show the method yields large, positive correlation coefficients for well pairs identified by tracer tests to have strong connectivity, and low positive, or negative coefficients for well pairs with known poor connectivity. The authors also showed that the data set being transformed must have sufficient temporal "texture" for wavelet analysis to be useful.

Some additional quantitative analysis has been done using synthetic tracer tests. One reservoir management concern is to identify the velocity of thermal fronts in the reservoir. The velocity of a temperature front, vT, is related to the fluid velocity, vw, in a fixed manner.[20]

Vol6 page 0418 eq 002.png....................(3)

By transforming tracer production data at each production well, Shook[1] showed that thermal velocities can be predicted from tracer tests. These studies were restricted to heterogeneous, nonfractured media and single-phase conditions, where thermal conduction is largely a secondorder effect. Efforts to extend the method to fractured media have met with limited success, in particular because of fracture geometry. Likewise, quantitative analysis of tracers in two-phase or superheated steam reservoirs is difficult. Because the tracer is transported by either of the phases at various times (e.g., vaporizing here and condensing there), mean residence times are more difficult to interpret. Under certain conditions, a boiling interface may develop between the fluid originally in place and the cooler injectate.[21] The velocity of this boiling front has been studied analytically,[22][23] and can be predicted for simple geometries and homogeneous reservoir conditions. In cases where buoyancy is important, however, the vaporized tracer may not trace injectate flow paths, making the interpretation still more difficult. Predicting thermal velocities in fractured media remains an active research topic in geothermal tracing.

Analysis of tracer tests conducted in geothermal fields ranges from purely qualitative to quantitative, volumetric analysis of pore volume. Matsunaga et al.[24] show an analysis of seven tracer tests conducted at the Hijiori, Japan, engineered geothermal system. By comparing mean residence times[25] for consecutive tracer tests, they showed that the flow system was evolving during the injection of cool (25 to 50°C) liquid into initially hot (~150°C) dry rock. They concluded in part that anhydrite scaling was plugging some of the fractures, thereby modifying the flow field. They also noted a rapid decline in produced temperature during the injection tests, but did not correlate the thermal velocities with tracer velocities. The Hijiori geothermal reservoir is among the most instrumented and studied engineered geothermal systems in the world. A variety of tracer tests have been conducted and reported on over a number of years.[24][26][27]

Other than the analyses for the Hijiori field tracer tests, a majority of tracer test interpretations remains qualitative. Fig. 1 is an example of analysis of several tracer tests conducted at Dixie Valley, Nevada. The geothermal field has been a test facility for testing naphthalene sulfonates for a number of years, and seven such tests have been conducted since 1997.[11] The relative size of the arrows is indicative of the relative contribution of an injector on a set of producers. Estimates of reservoir pore volume have also been calculated on the basis of tracer dilution.[28] However, the interpretation in the figure (i.e., relative contribution of injectors to production areas) remains the most-used information from these tracer tests.

An example of tracer test interpretation in vapor-dominated reservoirs is given in Fig. 2. This figure summarizes the interpretation of a tracer test conducted in The Geysers geothermal field in Northern California. In this test, two hydrofluorocarbons, R23 and R134a, and tritiated water were injected into a zone containing moderately (~15°C) superheated steam. Fig. 2 shows the cumulative mass fraction of R134a and tritium recovered from wells surrounding the injector. Tritiated water is a nearly ideal geothermal tracer because its properties are nearly identical with those of water and, therefore, tracks the injectate very well. Adams et al.[12] suggest that the similarity in recovery between the tritium and R134a suggests both compounds remained with the injectate, indicating R134a is a useful tracer for areas with low or moderate superheat. Another tracer test conducted in a highly superheated zone at The Geysers showed substantial separation between tritiated water and the chlorofluorocarbon R13.[12] The authors concluded that a large degree of superheat exaggerates the effect of volatility, and caution should be exercised in using tracers whose volatility greatly exceeds that of water when superheated conditions prevail.

While some tracer tests have been modeled,[29] this is one aspect of tracer test analysis that has tended to lag behind oilfield practices. Recent advances have been made in improving the phase behavior routines for vapor-liquid partitioning tracers,[3][30] and use of modeling tracer tests is expected to increase.

Nomenclature

ai = correlation coefficients between a given wells’ produced chloride and the ith injection wells’ injected chloride concentration
ao = initial chloride concentration for a given production well
cij = arc cost function for the travel arc between a given injection and production well pair (e.g., along a streamline)
cr = rock compressibility
cw = liquid compressibility
Clp = produced chloride for a given well
d = wellbore diameter
i = injector
qi = chloride injection rate in well i
qri = re-injection rate at well i
vT = velocity of temperature front
vw = fluid velocity
W = mass flow rate
ρ (p,T) = fluid density
ρr, ρw, ρv = densities of rock, water, and steam, respectively
Φ = porosity

References

  1. 1.0 1.1 Shook, G.M. 2001. Predicting Thermal Breakthrough in Heterogeneous Media from Tracer Tests. Geothermics 30 (6): 573. http://dx.doi.org/10.1016/S0375-6505(01)00015-3
  2. 2.0 2.1 Rose, P.E. et al. 1997. Numerical Simulation of a Tracer Test at Dixie Valley, Nevada. Proc., Twenty- Second Workshop on Geothermal Reservoir Engineering, Stanford U., Stanford, California, 169.
  3. 3.0 3.1 Bloomfield, K.K. and Moore, J.N. 2002. Modeling Hydrofluorocarbon Compounds as Geothermal Tracers and Design of a Two-Phase Tracer Test. Geothermics 32 (3): 203. http://dx.doi.org/10.1016/S0375-6505(03)00017-8
  4. 4.0 4.1 4.2 Sullera, M.M. and Horne, R.N. 2001. Inferring Injection Returns from Chloride Monitoring Data. Geothermics 30 (6): 519. http://dx.doi.org/10.1016/S0375-6505(01)00016-5
  5. Beall, J.J. 1993. NH3 as a Natural Tracer for Injected Condensate. Geothermal Resources Council Trans. 17: 215.
  6. Nuti, S., Calore, C., and Noto, P. 1981. Use of Environmental Isotopes as Natural Tracers in a Reinjection Experiment at Larderello. Proc., Seventh Workshop on Geothermal Reservoir Engineering, Stanford University, Stanford, California, 85.
  7. Beall, J.J., Enedy, S., and Box, W.T. Jr. 1989. Recovery of Injected Condensate as Steam in the South Geysers Field. Geothermal Resources Council Trans. 13: 351.
  8. Gambill, D.T. 1992. The Recovery of Injected Water as Steam at The Geysers. Geothermal Resources Council Special Report 17: 159.
  9. 9.0 9.1 Gulati, M.S., Lipman, S.C., and Strobel, C.J. 1978. Tritium Tracer Survey at The Geysers. Geothermal Resources Council Trans. 2: 227.
  10. Adams, M.C. et al. 1992. Thermal Stabilities of Aromatic Acids as Geothermal Tracers. Geothermics 21 (3): 323. http://dx.doi.org/10.1016/0375-6505(92)90085-N
  11. 11.0 11.1 Rose, P.E., Benoit, W.R., and Kilbourn, P.M. 2001. The Application of the Polyaromatic Sulfonates at Tracers in Geothermal Reservoirs. Geothermics 30 (6): 617. http://dx.doi.org/10.1016/S0375-6505(01)00024-4
  12. 12.0 12.1 12.2 12.3 12.4 Adams, M.C. et al. 2001. Hydrofluorocarbons as Geothermal Vapor-Phase Tracers. Geothermics 30 (6): 747. http://dx.doi.org/10.1016/S0375-6505(01)00027-X
  13. Maxfield, B.T. et al. 2002. Evaluation of Fluorocarbon Tracer Retention in Dry and Wet Sand Column Tests. Geothermics Trans. 841–846.
  14. McCabe, W.J., Barry, B.J., and Manning, M.R. 1981. Radioactive Tracers in Geothermal Underground Water Flow Studies. Geothermics 12 (2–3): 83. http://dx.doi.org/10.1016/0375-6505(83)90020-2
  15. Giovannoni, A. et al. 1981. First Results of a Reinjection Experiment at Larderello. Proc., Seventh Workshop on Geothermal Reservoir Engineering, Stanford U., Stanford, California, 77.
  16. Horne, R.N. 1982. Effects of Water Injection into Fractured Geothermal Reservoirs—A Summary of Experience Worldwide. Geothermal Resources Council Special Report 45.
  17. 17.0 17.1 Lovekin, J. and Horne, R.N. 1987. Optimization of Injection Scheduling in Geothermal Fields. Geothermal Resources Council Trans. 11: 607.
  18. Harper, R.T. and Jordan, O.T. 1985. Geochemical Changes in Response to Production and Reinjection for Palinpinon Geothermal Field, Negros Oriental, Philippines. Proc., New Zealand Geothermal Workshop, Auckland, New Zealand, 7, 39–44.
  19. Macario, E.G. 1991. Optimizing Reinjection Strategy in Palinpinon, Philippines Based on Chloride Data. MS thesis, Stanford University, Stanford, California.
  20. Bodvarsson, G. 1972. Thermal Problems in Siting of Reinjection Wells. Geothermics 1 (2): 63. http://dx.doi.org/10.1016/0375-6505(72)90013-2
  21. Shook, G.M. 2001. Thermal Velocities Arising from Injection in Two-Phase and Superheated Reservoirs. Proc., Twenty-Sixth Workshop on Geothermal Reservoir Engineering, Stanford U., Stanford, California, 197.
  22. Pruess, K. et al. 1987. An Analytical Solution for Heat Transfer at a Boiling Front Moving through a Porous Medium. J. of Heat & Mass Transfer 30 (12): 2592. http://dx.doi.org/10.1016/0017-9310(87)90140-2
  23. Woods, A.W. and Fitzgerald, S.D. 1993. The Vaporization of a Liquid Front Moving through a Hot Porous Rock. J. of Fluid Mech. 251: 563. http://dx.doi.org/10.1017/S0022112093003520.
  24. 24.0 24.1 Matsunaga, I. et al. 2002. Reservoir Monitoring by Tracer Testing During a Long-Term Circulation Test at the Hijiori HDR Site. Proc., Twenty-Seventh Workshop on Geothermal Reservoir Engineering, Stanford U., Stanford, California, 101.
  25. Levenspiel, O. 1972. Nonideal Flow. Chemical Reaction Engineering, Ch. 9. New York City: John Wiley & Sons, Inc.
  26. Matsunaga, I., Tao, H., and Kimura, A. 1996. Preliminary Characterization of the Hijiori HRD Deeper System by Fluid Geochemistry and Tracer Experiments of a One-Month Circulation Test. Proc., Third International HDR Forum, Santa Fe, New Mexico, 25–26.
  27. Oikawa, Y. et al. 2002. Heat Extraction Experiment at Hijiori Test Site. Proc., Twenty-Seventh Workshop on Geothermal Reservoir Engineering, Stanford U., Stanford, California, 89.
  28. Rose, P.E., Apperson, K.D., and Faulder, D.D. 1997. Fluid Volume and Flow Constraints for a Hydrothermal System at Beowawe, Nevada. Presented at the SPE Annual Technical Conference and Exhibition, San Antonio, Texas, 5-8 October 1997. SPE-38762-MS. http://dx.doi.org/10.2118/38762-MS
  29. Birdsell, S. and Robinson, B. 1988. A Three-Dimensional Model of Fluid, Heat, and Tracer Transport in the Fenton Hill Hot Dry Rock Reservoir. Proc., Thirteenth Workshop on Geothermal Reservoir Engineering, Stanford University, Stanford, California, 225.
  30. Trew, M., O’Sullivan, M.J., and Yasuda, Y. 2001. Modeling the Phase Partitioning Behavior of Gas Tracers under Geothermal Reservoir Conditions. Geothermics 30 (6): 655. http://dx.doi.org/10.1016/S0375-6505(01)00025-6

Noteworthy papers in OnePetro

Use this section to list papers in OnePetro that a reader who wants to learn more should definitely read

External links

Use this section to provide links to relevant material on websites other than PetroWiki and OnePetro

See also

Geothermal reservoir engineering

Geothermal reservoir characterization

Modeling geothermal reservoirs

Geothermal drilling and completion

Geothermal energy

PEH:Geothermal Engineering