Cyclic Steam Stimulation Results in High Water Retention for Kuwaiti Heavy-Oil Field

Fig. 1—Temperature survey of Well 1.

You have access to this full article to experience the outstanding content available to SPE members and JPT subscribers.

To ensure continued access to JPT's content, please Sign In, JOIN SPE, or Subscribe to JPT

Cyclic steam stimulation (CSS) is one of the principal enhanced-oil-recovery methods for heavy oil. CSS was performed in some of the wells of a heavy-oil field in Kuwait. Multiple cycles were applied in these wells. However, the total water produced in each cycle was much less than expected. This paper presents experiments that were conducted to find possible reasons for the high water retention.

Hysteresis in CSS

Hysteresis in drainage and imbibition nonwetting-phase capillary pressure and relative permeability curves is an established phenomenon. Hysteresis of capillary pressure and relative permeability to water has a great effect on heavy-oil recovery and producing water/oil ratios (WORs) during cyclic steam stimulation. Produced WOR calculated without considering hysteresis will be unrealistically high relative to that observed in the fields.

Water Retention

During water injection, relative permeability to water increases, and, during the production phase, it decreases; however, it does not follow the same path. Relative permeability to water during the production phase is always less than that during the injection phase at the same water saturation. Irreducible water saturation during the production phase is more than the initial water saturation before injection. The difference between these two water saturations is water retention.

The same phenomenon occurs when CSS is applied. As a result of hysteresis and temperature, irreducible water saturation does not reach the initial water saturation. When steam is injected into a formation, the formation becomes increasingly water-wet. In subsequent cycles, steam injection reduces the oil saturation and increases the water saturation around the wellbore. Relative permeability to oil at high water saturation is also reduced. In later cycles, increasingly more water is produced than oil, and, therefore, the WOR increases. The produced water consists of both the condensed steam and the formation water. After steam is injected, fresh water is produced first, and slowly the salinity of the produced water increases. Initially, some of the condensed steam returns, but, later, formation water is produced with the condensed steam. If the well produces for long enough, eventually only formation water will be produced.

Field Under Consideration

The Kuwaiti field considered in this study is spread over 1200 km2, consisting of a large heavy-oil accumulation in a shallow reservoir. The formation depth varies from 400 to 850 ft. The formation is divided into four major sand units—Upper A, Upper B, Lower A, and Lower B, from top to bottom, where a marine cap shale exists on top of the formation. A cemented siltstone exists between the cap shale and Upper A, and a silty cemented baffle separates Upper A and Upper B. The initial understanding was that Upper A and Upper B were separated by a regional shale barrier, but studies with increased certainty reveal that the so-called shale barrier is mainly composed of cemented sandy siltstone (with minor mudstone), which is primarily a baffle unit with a permeability between 15 and 350 md. Lower A and Lower B are also separated by a baffle. Another quite shaly baffle exists between the upper and lower sand units in the northern part of the field and may act as a barrier.

Performance of CSS Wells

The injection volume and production times varied greatly from well to well and from cycle to cycle. The steam-slug sizes varied from 204 to 904 bbl/ft net pay in the first cycle, from 291 to 881 bbl/ft net pay in the second cycle, and from 229 to 242 bbl of cold-water-equivalent (CWE) steam per foot of net pay in the third cycle. The production period varied from 215 to 946 days in the first cycle, from 246 to 749 days in the second cycle, and from 427 to 649 days in the third cycle. Water retention varied from –246 to 68%, and produced WOR was between 0.08 and 0.42 for the entire period of the first cycle. Because the production phase was long, the water retention and produced WOR for the first 200 days of production were calculated considering that the production phase should be approximately 200 days in each cycle. Water retention varied from 27 to 74%, and produced WOR was between 0.13 and 0.74 for the first 200 days of production. The produced water contained condensed steam and some formation water. Water retention is calculated by this relation: % Water Retention=(CWE Steam Injected−Water Produced)×100/CWE Steam Injected.

Laboratory Investigation of Water Retention

Two steamflood experiments were conducted to determine water retention. Composite samples consisted of five 2-in. plugs for the first experiment and six plugs of the same size for the second ­experiment. The initial water saturation was measured before the steamflood. Then, 100%-quality steam was injected at 400°F and a constant rate of 4 cm3/min.

This continued until a 99.9% water cut was achieved. Following the steamflood test, the composite sample was prepared for drainage measurements. The temperature was set at 150°F, and oil was injected from the steamflood production end to mimic the CSS process. Oil was injected until a producing oil cut of 99.9% was achieved. The final irreducible water saturation then was determined. The difference between the final irreducible water saturation and initial irreducible water saturation was calculated to find the amount of water retained because of hysteresis and temperature.

In these two experiments, water retention was found to be 19 and 12%, respectively. The experiments suggest that water retention in the field should be much lower than what was observed.

Discussion of Water Retention in the Formation

Four cycles were applied in one well, three cycles in another, two cycles in four more wells, and one cycle in a final well. Water retention varied between 27 and 68% in that first cycle, between –107 and 84% in the second cycle, and between –73 and –35% in the third cycle. Likewise, the produced WOR varied between 0.13 and 0.38 in the first cycle, between 0.22 and 0.99 in the second cycle, and between 0.36 and 0.44 in the third cycle. In all but one well, the produced WOR increased with each cycle in the first 200 days of production. Water produced from these wells was much less than expected. During the entire long period of production in the first cycle, produced WOR was in the range of 0.08 to 0.42, which is quite low.

In the first cycle, the average water retention was 55%. In the second cycle, water retention was also too high and all but one well did not produce water equal to the injected volume. Overall, water retention was high and produced WOR was low in all cycles. In none of the cycles did WOR even reach unity. The production figures and salinity determined during the production period suggest that injected steam did not come back fully even after three cycles. Two wells that underwent two cycles performed better comparatively in terms of water retention and produced more water than injected volumes in both cycles. Both of these wells are completed in Upper A where there is a comparatively thin baffle above the oil formation. Cap shale exists above the baffle zone and does not allow the steam to pass through. However, the produced WOR was lower than expected. It appears that the injected steam might have gone into a thief zone that has not yet been identified.

A temperature survey was conducted in all of the wells. The temperature surveys of Well 1 and Well 6  in the Upper B are shown in Fig. 1 above and Fig. 2, respectively. The figures show that the temperature increased in the baffle zone just above the sand formation. The same phenomenon also was observed in other wells. In Well 1, the temperature of the baffle zone was almost equal to that of the sand formation into which the steam was injected. This increase in temperature could not result from conduction only, meaning that the injected steam entered the baffle through convection and raised its temperature. Therefore, the baffle acted as a thief zone.

Fig. 2—Temperature survey of Well 6.


The permeability of the baffle zone falls between 15 and 350 md, but the baffle has high water-saturation values. The baffle can impede the upward flow under normal conditions but cannot stop the flow completely. If formation pressure or temperature increases, the baffle becomes more active. During CSS, a high volume of steam was applied to the wells at high pressure. At that pressure, the baffle zone becomes more active.

With the knowledge that steam entered the upper baffle, the challenge is to discover why it did not return during production. A minimum differential/entry pressure is required for the steam to enter the baffle. This minimum pressure was achieved when the steam was applied. During the production phase, however, the reservoir pressure decreased, creating some drawdown but not enough to produce the condensed water back from the baffle.

This article, written by Special Publications Editor Adam Wilson, contains highlights of paper SPE 184154, “High Water Retention in Cyclic-Steam-Stimulation Wells of Kuwait Heavy-Oil Field,” by Anil Kumar Jain, Amal Al-Sane, and Fatma Ahmad, Kuwait Oil Company, prepared for the 2016 SPE International Heavy Oil Conference and Exhibition, Mangaf, Kuwait, 6–8 December. The paper has not been peer reviewed.

Cyclic Steam Stimulation Results in High Water Retention for Kuwaiti Heavy-Oil Field

01 March 2018

Volume: 70 | Issue: 3


Don't miss the latest content delivered to your email box weekly. Sign up for the JPT newsletter.