The major productive interval of the field, includes a thick section composed of moderate-to-high permeability fluvial sands and interbedded shales. Although the upper sand is perforated (and often hydrualically fractured) in both injection and production wells in some areas, it contributes relatively little to the injection capacity and receives very little injection. An extensive non-pay heavy oil/tar zone (HOT) underlies the oil column and prevents aquifer influx in most of this area.

The upper sand is typically 1-10 md, Zone 4 varies from 20-175 md, and Zone 3 is 400 md or so. Permeability tends to decrease upwards within layers. Figure 1 is a schematic. The upper sand does not appear in this view. It can be seen from the gamma log trace for one of the wells analysed (Figure 2).

Figure 1. Generic lower reservoir geometry. Notice that the upper sand is not shown. Refer to Figure 2.

Figure 2. Example gamma ray trace, showing the upper sand from approximately 9610 to 9650 feet. Injection profiles for produced water and seawater are superimposed, showing essentially no flow in the upper sand for either fluid type.

1. In most instances in this scenario, thermal effects were the overriding mechanism. Specifically, the change in stress governed fracture vertical extent. Many of the comparative observations between seawater and produced water can be strongly linked to thermal stress changes. The difficult issue is to infer to what extent thermoelastic effects versus plugging govern the temporal development of the injection profile.
2. There is a thick, lower sequence of sands and an overlying siltier sand, separated from the lower sands by a moderately thick barrier. In almost all circumstances it was difficult to partition significant sustained injection into the upper zone, even when it was explicitly hydraulically fractured. It cannot be exclusively stated from the data whether this was due to lithology or stresses. As indicated, the sands in this upper zone were usually siltier or shalier and have lower permeability, and/or the stresses may be higher in that zone.
3. A subjective overall conclusion is that the injection under fracturing conditions in this reservoir would seem to favor the higher permeability zones. The magnitude of the permeability is inferred subjectively from the gamma ray.
4. In a number of cases, the worst enemy to effective injection partitioning was poor cementing.
5. Figure 3 shows a difficulty that was encountered in a number of the wells. Whether it is physical (crossflow) or operational is difficult to determine. If it is operational, it emphasizes the need for appropriate calibration (this is taken as a given, but operational issues can always be difficult - low flow rates may be below spinner threshold levels, devices may be impaired by material, such as "schmoo," in the wellbore etc.). Ensure that calibration is performed, that tools with adequate spinner sensitivity are used and that material in the wellbore and plated against the formation/perforations does not impede tool operation.

   

   Figure 3. Gamma ray and PLT profiles, showing that the zones taking produced water progressively becomes restricted and higher up in the fracture with time. More importantly from an operational point of view is understanding/eliminating non-zero initial or final percentages (appropriate calibration - why is the initial percentage only 20% with no perforations below and/or why is the total percentage less than 100% with no perforations above), and/or from a physical point of view, resolving and determining if crossflow is legitimately detected. The appearance of the lower part of the produced water profile from March 1, 1992 either suggests crossflow or that the lower set of perforations are not taking fluid (fill can be ruled out because the tool was passed to at least 11175 feet.

6. In many cases, it was not possible to determine lithologically controlling mechanisms in the lower sands. They have approximately the same gamma ray signature and this was the only information available.
7. In almost all cases, the produced water profile was somewhat different than the seawater profile. In a reasonable number of cases, injectivity into the lowermost parts of the lower sands was not as effective for produced water. Reasons for this can include fill in the wellbore, fracture closure due to stress levels, and possibly preferential blocking in the fracture. As Murray (circa 1994) observed when he evaluated the data:
   • "Well conformance is better on SWI than on PWRI, possibly due to plugging of the perfs and filling of the sump by PW sludge [supported by current evaluations]. The greatest fraction of flow tends to be taken by the highest permeability intervals [supported by current evaluations]. On average, the height of the main zone of flow on PW is smaller than on SW, possibly due to fracture related effects [the current evaluation also indicated that the primary depth range of injection was somewhat different for seawater and produced water, often moving up with produced water - again suggesting damage, closure or fill]."
   • "The injection of MI gas has a slightly beneficial effect on short term PW injectivity, but it impairs short term SW injectivity. The most likely reason for this behavior is the change in fracture size due to the formation temperature change on MI, since MI gas is slightly cooler than PW but significantly warmer than SW." This is an important observation. Note also that MI was under matrix conditions [using the strict definition of matrix injection - i.e., there may have been aq conductivie, self-propped fracture, but it was stationary] and this can be used to explain those situations where there was less coverage than for the produced water. As pointed out by Murray: "it is probable that MI injection is also governed by the presence of only partially healed fractures."

1. In general, the PLT information was good quality. There are some apparent calibration or spinner threshold issues. This is not too surprising with the fill in these wells.
2. While pressure data from the PLT surveys were almost certainly taken, they were not available for this data set and would have been quite difficult to ressurect. The pressure data would have been valuable for pressure transient analysis.
3. While temperature survey information was available on a number of the wells, it was characteristically much more difficult to interpret than spinner data. Electronic temperature data were not available. The message might be to consider carefully the value of temperature data (and there can be value), be careful about misinterpretation (there have been cases where channel flow has mistakenly been interpreted as longitudinal fracturing) and when possible use differential temperature interpretation (comparing subsequent temperature surveys, taken at different times on the same survey, with adequate time allowed for cooldown or heat up).
• While the decision is ultimately economic, it is commonly much easier to mitigate than to remediate. Once nonuniform profiles develop, or excessive reservoir crossflow occurs, it may be difficult to overcome it effectively.
4. Be careful (and creative) on how you bring your injectors on line
• Methods are presented for minimizing the impact of nonuniform profiles by drilling, by completing in particular manners, by completing effectively, with isolation and by manipulating production. Don't forget to use the natural tendencies of formations to experience stress changes due to temperature and formation pressure changes. There are a number of documented cases where thermal stress manipulation during beanup has led to significant improvements in flow distribution that would not otherwise have been possible.

Document	Description	Download Size
Real Data (Summary)	Certainly this is a situation that is common. However, there is often not a substantial amount of measurement data. One exception is for a rich set of information from operations on the North Slope. Download a summary.	784 kB
Real Data (Details)	Certainly this is a situation that is common. However, there is often not a substantial amount of measurement data. One exception is for a rich set of information from operations on the North Slope. Download more details including tables of operational events, gamma logs, PLT information, etc.	2.31 MB
Real Data (All)	Certainly this is a situation that is common. However, there is often not a substantial amount of measurement data. One exception is for a rich set of information from operations on the North Slope. Download all of the above plus sanitized Excel files for the PLT and injection data.	31.319 MB