Cross-Well EM Images Fluid Dynamics

FEBRUARY 2010 The Better Business Publication Serving the Exploration / Drilling / Production Industry Cross-Well EM Images Fluid Dynamics By Michael Morea, Ajay Nalonnil and Michael Wilt BAKERSFIELD, CA. Accurately delineating reservoir fluid dynamics has long been a dream for the oil and gas industry. Improving this capability is more urgent than ever, especially in mature oil provinces such as the United States, where the bulk of production still comes from fields that are decades old. A lack of access to some of the biggest undrilled U.S. prospects has resulted in a lagging oil discovery rate, and it has fallen to secondary and tertiary recovery projects to keep the nation s oil production from declining. The industry s extensive efforts to sustain domestic production through optimized secondary and tertiary oil recovery techniques have helped to keep U.S. oil production and reserves from falling more precipitously than they would have under typical oil field primary decline rates. Indeed, enhanced oil recovery projects now account for 12 percent of total U.S. The magnetic moment produced by the cross-well electromagnetic transmitter is 100,000 times stronger than the source in a conventional single-well induction logging tool, generating res - ervoir-scale imaging results in open- and cased-hole wells that are comparable to CAT scan imaging technology. oil production. The vast majority of domestic oil fields have undergone or are undergoing waterflooding. Every bypassed pocket of oil and every unswept zone becomes a crucial consideration for operators. Therefore, it is critical for operators to obtain the most detailed picture of reservoir fluid dynamics possible, and at a reservoir scale. Well bore logging and seismic are among the most widely used tools to delineate fluid behavior in reservoirs. However, both have their shortcomings. Bore hole induction logging provides excellent detail to within a few feet of the reservoir surrounding the well bore wall, but conventional induction logging falls short when it comes to characterizing reservoir fluid behavior at the reservoir scale. Conversely, conventional seismic data acquisition provides coarse detail at the reservoir scale, but not the kind of fine resolution needed to determine fluid properties in a reservoir. A new innovative tool has been developed to bridge this gap: electromagnetic (EM) enhanced cross-well reservoir imaging and monitoring system. The technology measures resistivity at a reservoir scale between well bores that can be as much as 3,280 feet apart. Measuring resistivity sensitivity to changes in fluid saturation and temperature makes this cross-well EM system ideal for tracking the distribution of injected fluid volumes and the resulting swept zone. With it, an operator can now image and monitor the effect of steam or water saturation changes, which aids in guiding field development and enhancing reserve estimates. The system employs electromagnetic physics akin to that in conventional in- Reproduced for Schlumberger with permission from The American Oil & Gas Reporter

FIGURE 1 SF LA SD duction logging. However, electromagnetic enhanced cross-well reservoir monitoring entails deploying a dynamic transmitter sonde in one well and a receiver array in an offset well. The wells may be hundreds or even several thousands of feet apart. With cross-well EM technology, the magnetic moment produced by the transmitter is 100,000 times stronger than the source in a conventional single-well induction logging tool. The imaging result is comparable to that obtained with a CAT scan. The system can work with either open-hole or cased wells. Ideally, electromagnetic enhanced cross-well reservoir monitoring is used in a time-lapse mode to accurately track the movement of fluids over time. For example, applying time-lapsed EM enhanced cross-well reservoir monitoring to a water-alternating-gas injection program allows the operator to assess, over time, water movement through a reservoir and the ideal injection profile for increasing oil recovery, while avoiding problems such as water override. The system provides a real-time assessment of sweep effectiveness and also can help identify bypassed pay. Cymric Field Location Map Open Hole Crosswell Tomography Survey Area California Case Study An early example of a successful application of a cross-well EM survey in the United States involved a Chevronoperated project in the Cymric Field in California s San Joaquin Valley. Cymric, which celebrates its discovery centennial this year, produces mainly heavy oil from several reservoirs using cyclic steam drive. The Cymric 1Y reservoir consists of the Antelope Shale member of the Miocene Monterey formation. More than 400 wells have been drilled into this shallow, 600- foot-thick reservoir. The Cymric 1Y reservoir structure is a fairly simple southeast-plunging anticline with steeply dipping beds (Figure 1). The top of the reservoir is marked with a dipping unconformity, and the main reservoir dips at 40-60 degrees. The complex, steeply dipping geology of the Cymric 1Y reservoir makes it difficult to gauge the affected volume around each cyclic injector, and the vertical and radial distribution of the steam from the well. The challenge was to determine the steam-saturated volume around cyclic injector wells in relation to the local geology in the Cymric 1Y reservoir. The goal of this application was to use Schlumberger s DeepLook-EM cross-well imaging system deployed in a single well pair to identify the steam front based on changes in fluid saturation and temperature. Specifically, the survey was designed to evaluate the feasibility of applying interwell EM to image siliceous shale stratigraphy and steaminduced fractures from cyclic steam development of the Cymric 1Y. FIGURE 2 Sample Cymric Resistivity Logs (Before and After Steaming) Original Well Swept Well

Chevron uses a cyclic steam drive at Cymric 1Y, wherein steam is injected at pressures exceeding fracture gradient to create a steam-induced fracture. While this process has proven successful, there were several questions that had to be addressed to optimize field development, ultimate oil recovery and reserves estimates, including: What is the steam volume affected around each cyclic injector? How is the steam distributed vertically and radially from the well? Does the steam volume change with each cycle? How does the local geology influence the steam-swept zone? FIGURE 4 Depth (ft) Cross-Well Resistivity Section with Stratigraphic Markers Superimposed T01303A T01304 13041 Lab And Field Testing Cross-well EM measures formation resistivity between wells. These measurements make it possible to infer structure, temperature distribution, and residual saturation of affected reservoir volumes. A laboratory experiment with a water-saturated sand core showed that the formation resistivity decreases by more than a factor of five as the temperature increases from ambient to 200 degrees Celsius (392 degrees Fahrenheit). This has been shown to be the dominant effect in many steamfloods and affirms interwell resistivity as an excellent indicator of the formation temperature. Another lab test, this time with a saltwater-saturated sand core, showed an even more striking relationship between formation resistivity and water saturation. Extrapolating these results to the oil phase in a steam reservoir suggests that saturation effects in steamfloods can be very large. FIGURE 3 Cross-Well Resistivity Section Between Well Nos. 1303A and 1304 Depth (ft) T01303A T01304 13041 150 125 100 75 50 25 0 0.0 Location (ft) 0.5 1.0 1.5 Log Resistivity 2.0 150 125 100 75 50 25 0 0.0 Location (ft) 0.5 1.0 1.5 Log Resistivity 1,500 2.0 1,500 In effect, cyclic steaming at Cymric displaces oil with hot water and steam. It follows that resistivity effects would reflect both temperature and saturation conditions. This also implies that the temperature must be known or estimated to obtain a saturation distribution from the cross-well resistivity section. Figure 2 is a site map of part of Cymric 1Y that encompasses the wells used in the survey. Well No. 1304I is a cyclic well drilled in 1987 and abandoned in 2000. Well No. TO1304 is an observation well drilled in May 2003 and located within 50 feet of No. 1304I. The latter was drilled to assess the effect of cyclic steam injection on formation properties, as well as for future temperature monitoring. This well was extensively logged and cored to ascertain the effect of the injected steam on the reservoir. Figure 2 also shows formation resistivity logs from wells 1304I (prior to steaming) and TO1304. The logs show that the high-resistivity, oil-saturated intervals are replaced with low-resistivity water and steam-saturated rock. Since Well TO1304 encountered temperatures up to 250 degrees greater than those in 1304I (presteam ambient reservoir temperatures), it must be concluded that some of this change in resistivity is caused by temperature. However, most of the effect is likely caused by the variation in

saturation. The cross-well EM technology was deployed in these two wells to map the steam-affected zone in the interwell space. The goal was to delineate the boundaries of the affected reservoir and then determine if any identifiable structure was influencing steam flow. Data Collection, Processing The cross-well EM survey data were collected immediately after other openhole geophysical logs were run in Well TO1304 and prior to setting steel casing. The transmitter sonde was deployed in open-hole TO1304, and the four-level receiver tool was deployed in the subsequently drilled cyclic well, No. 1303A. The transmitter covered a depth range of 850-1,600 feet at feet an hour while continuously transmitting a sinusoidal signal. The signal was detected in Well 1303A, using receivers that remained fixed at four depths spanning a 60-foot interval. The receivers were then repositioned, and the transmitter traversed the same depth range again. This continued until both transmitter and receivers occupied all required positions in both wells targeted intervals. After data collection, the profiles were reduced. Data processing initially involved a data editing program that removed spurious points, calibrated the data, and resampled the data to a regular grid. Then, proprietary software was used to adjust the data for the effect of steel casing in Well 1303A. The data were then input into a 2-D automatic inversion code for interpretation. That entailed first resampling the data so that the source and receivers fell on a regular grid. This sped the process for computer inversion and resulted in no loss of accuracy because the profiles were sampled so densely. The next step was to use the original resistivity logs from each well, interpolated into the interwell region, to provide a starting model for the 2-D inversions. After horizontal smoothing was employed to emphasize layered structures, the computer then modified this starting model to minimize the fit between calculated data and observed data to within a specified tolerance, usually 1-2 percent. Finally, 10-15 iterations were required on a fast personal computer workstation to achieve an acceptable result. The computer also tabulated any misfits and tracked formation resistivity changes to the input model during the inversion. The Cymric 1Y inversion required several starting models before reaching a stable solution. Results And Observations Figure 3 shows the 1303A-TO1304 resistivity cross-section as a color image. Cooler colors (blues) denote the lowerresistivity characteristic of steam-swept zones, and the warmer colors indicate the higher-resistivity characteristic typical of unswept zones. The interwell resistivities were plotted on a logarithmic scale, and the color-coded resistivity logs are shown at the margin of the plot. The logarithmic scale was required because of the large resistivity contrast within the section, and the large change before and after steam sweeping. The contrast between the resistivity logs of the original producer (1304I) and the nearby observation well (TO1304) is striking. The oil-rich horizons have declined in resistivity from more than 50 ohm-meter to less than 2 ohm-meter, in some cases. This was most likely a saturation effect, since the temperature could account for a relatively modest percentage of this change. Most of the oil horizons are fully swept. The cross-section shows an abrupt boundary located about 75 feet from Well TO1304, where the resistivity changes from 2 ohm-meter to more than 50 ohmmeter over a short interval. The lower resistivity is associated with the depleted zone, as evidenced by the log from TO1304. The low resistivity is predominantly the result of formation water and steam condensate displacing oil. The abrupt nature of the boundary suggests that the transition between swept and unswept formation is also abrupt. Figure 4 shows the same cross-section, but with stratigraphic markers superimposed. The dipping horizons indicated in Figure 4 put the resistivity cross-section in context and permit some observations. First, the boundary between the swept and unswept zones is an inclined structure roughly perpendicular to the formation dip. The lowresistivity zone seems to move from the bottom upward. The shallower interwell horizons are not fully swept. It is likely that the steam will naturally follow the high-permeability channels and migrate upward because of its buoyancy. Second, depletion may be dependent on geologic factors such as lithology, dip, silica phase, fractures, etc. The maximum extent of lateral depletion is close to 120 feet at the lower interfaces, but less in the shallower zones. The average extent is roughly 75 feet. The boundary at the lower zones is much less clear. In addition, although the resistivity in the cross-section roughly follows the superimposed dipping structure, there appears to be some horizontal smoothing MICHAEL MOREA is a senior staff geologist with Chevron in Bakersfield, Ca. He has been with Chevron since 1981. Morea s previous assignments with the company included Alaska and California exploration, regional stratigraphic studies, designing new enhanced oil recovery projects, property trades and acquisitions, and project management. Morea holds a Ph.D. in geology from the University of California-Riverside. AJAY NALONNIL is Schlumberger s product champion for DeepLook-EM. Based in Richmond, Ca., he has 14 years of experience with Schlumberger, starting as a field geophysicist processing land and marine seismic data in the Asia Pacific region. Following that field work, Nalonnil served at Schlumberger Information Solutions in reservoir modeling workflows, focusing on seismic-to-simulation projects in Asia. He holds a B.S. in applied geophysics from Curtin University, and an M.B.A. in technology management from Deakin/La Trobe University in Australia. MICHAEL WILT is the business development manager for bore hole electromagnetics for Schlumberger, based in Abu Dhabi. Previously, he was with Electromagnetic Instruments Inc. (acquired by Schlumberger in 2001), where he was the leader of bore hole EM research efforts and helped develop the first commercial cross-well EM field service. Between 1989 and 1997, Wilt was employed at Lawrence Livermore National Laboratory, where he applied bore hole electrical and electromagnetic methods to oil and geothermal field characterization and steamflood monitoring. He also was employed as a staff scientist at Lawrence Berkeley National Laboratory, specializing in geothermal exploration technology. Wilt holds a B.S. and an M.S. in geophysics from the University of California, Riverside, and a Ph.D. from the University of California-Berkeley.

that crosses bedding, which is attributed to the processing. The cross-well EM survey imaged a reasonable interwell structure consistent both with the collected data and the known geology. It successfully identified the steam-saturated volume associated with the cyclic steam injection within the surveyed section. This bodes well for future surveys of this type in this area. Since the California tests, the crosswell electromagnetic system has been deployed in projects in the Middle East, following extensive efforts to improve data quality by increasing frequency, slowing logging speed, and improving signal-to-noise results. This has resulted in even higher-resolution models. Cross-well EM has been shown to be a very promising technology for monitoring sweep efficiency, detecting bypassed pay, and optimizing reservoir simulation. Of special value is its capability for telescoping the timeline for tracking and monitoring reservoir fluid movements to a matter of months, rather than years. By monitoring the movement of fluids at the reservoir scale, cross-well EM can help operators understand reservoir dynamics more precisely to improve efforts to predict reservoir behavior. r