Before its merger with PetroFina, nearly half Total's worldwide reserves and 38% of its production came from the Middle East. Those percentages may continue, if the company secures further involve ment in major projects offshore Iran and Iraq.
The United Arab Emirates offer more modest growth opportunities, but they remain the cornerstone of Total's activities in the region. The company has stakes in some very long term production projects offshore Abu Dhabi and Dubai. Total became Abu Dhabi's first international offshore production operator in 1974, and to this day is the major international player in oil and gas in the UAE.
As a founder member, with BP, of the Abu Dhabi Marine Areas (ADMA) Operating Company, Total discovered the Umm Shaif Field in 1958 and Zakum Field in 1963. First oil from Umm Shaif was exported in 1962.
In 1974, newly created state oil company ADNOC assumed control of 60% of ADMA's oil operations as well as outright ownership of its gas interests. That has left Total today as a 13.33% partner in ADMA-OPCO, which operates Umm Shaif and Lower Zakum. These remain Abu Dhabi's largest producing offshore fields.
At around the same time, Total was granted operatorship of the smaller Abu Al Bukhoosh (ABK) field, discovered in 1968, 180 km offshore, and close to the median line with Iran. Japan's Inpex, with 25%, is the sole partner in this continuing development.
Over the ensuing decades, Total spearheaded the introduction of more advanced technologies for these fields, particularly ABK, to offset natural pressure decline. It has also assisted ADNOC in building its own technical expertise in upstream services.
Umm Shaif/Zakum
Umm Shaif, comprising several reservoirs, and Lower Zakum (the upper part of this field is operated by another company, ZADCO) generate 500,000 b/d combined. Potential of these fields remains so huge that the partners still plan in terms of 25-year production plateaus, which are effectively nominal.
"Not many oil companies in the world are in this position," says Jean-Francois Arrighi de Casanova, Total's Vice-President for UAE operations. "It's an ongoing situation - every year we plan ahead for the next 25 years, to ensure that the fields will be able to sustain capacity."
There is a network of remotely operated satellite and wellhead platforms on these fields linked by flowlines to two central production complexes, where first-stage separation is performed. "When natural depletion first occurred in the 1970s, we applied water injection. Currently, the fields are undergoing secondary recovery using water injection. A number of 100 MMcf/d gas injection pilots are also being trialed to assess the feasibility of implementing a larger-scale scheme."
Gas from these fields is also exploited in much larger quantities by ADGAS, the Persian Gulf's first LNG project. ADGAS is a consortium comprising ADNOC as 70% operator, Mitsui (15%), BP (10%), and Total (5%). Over 1,000 MMcf/d from 11 different sources is piped to the three-train liquefaction complex on Das Island, which has an annual capacity of 5.4 million tons. Much of this is transported by LNG tankers to Tokyo Electric Power in Japan, with the balance sold to customers in Europe, East Asia, and the US.
ABK
ABK - where Total holds the concession until 2018 - is a much more complex field to manage. To date, there have been seven separate phases of development, based around 22 fixed platforms in 30 meters water depth, with over 75 producer and injector wells and 38 sea lines. At peak, output was 80,000 b/d, but this has leveled off at around 30,000 b/d. Over 450 million bbl have been produced since startup in 1974.
Not long into the field's working life, ESPs had to be deployed to boost production, as pressure fell away in the field's multi-layered reservoirs. The quality of these reservoirs also varies considerably, according to Georges Buresi, Total's Vice President, Middle East. In addition to artificial lift, gas lift was employed using associated gas from ABK's crude. Then water injection was implemented in the 1980s for secondary recovery.
Two years ago, oil recovery via tertiary gas injection was also introduced. "So all these techniques are being implemented at the same time, in addition to some highly deviated multi-drain horizontal wells currently," says Buresi. The 1,358-meter-long horizontal AK-25B well, completed in 1994, intersects three oil-bearing horizons in the deep Araej reservoir. "We have been shooting 3D seismic with specifically adjusted reservoir imaging to reach various oil pockets, in order to maximize recoverable oil," he adds.
Associated gas production from the underlying Khuff reservoir started in 1994. Current output is 250 MMcf/d, with capacity on the dedicated processing platform for 300 MMcf/d. This level will be doubled in two years, with six more wells to be brought onstream, drilled from a duplicate platform currently under construction.
Tertiary scheme
Following two successful pilots, Total implemented ABK's first full-scale, non-miscible, tertiary gas injection scheme in a carbonate reservoir rapidly approaching depletion. The project was explained at the 8th Abu Dhabi International Petroleum Conference in October 1998.
ABK is located in the southern part of an anticline stretching across the UAE-Iran median line. The field's 160-meter-thick Arab D2 reservoir comprises eight layers with thicker limestone units (up to 25 meters) and typically thinner dolomite units (up to 8 meters).
The reservoir fluid is highly unsaturated, with initial pressure of around 3,900 psi and saturation pressure of 600 psi. It is produced under natural water drive via a strong aquifer. In 1995, following 21 years in production, 22 gas lift wells were producing the reservoir with average water cut of 86%.
Although recovery was good, the presence of a lean gas source led Total to investigate gas injection as an alternative. An earlier swelling test had shown that at reservoir conditions, 0.5 moles of gas per mole of oil could be dissolved, leading to a volume increase of 16% and decreased viscosity of 0.55 cP (from 1.07 cP).
A first pilot gas injection scheme in 1991 had proved disappointing, with gas breakthrough and oil gains observed only on one producer. Results were better from a second gas injector drilled in 1993 in the west of the field with shorter spacing from existing producers. Breakthrough was recorded on all surrounding producer wells, with oil gains three times greater than from the first pilot.
Simultaneously, laboratory experiments had been conducted to generate a reservoir simulation model and to assess microscopic displacement efficiency of residual oil by gas. The results were not encouraging, and a follow-up experiment involving gas flood after complete water flooding only boosted recovery by a further 4%. However, that outcome was put down to low permeability of the experimental core, which was not representative of the reservoir.
Full field study
Total derived enough satisfaction from these schemes to pursue a full-field study. According to the authors, a simulation model was built and matched to the pilot results, with inbuilt reservoir heterogeneity (high permeability streaks at the top and bottom of the D2 layer, and lateral barriers near the pilot 1 injector). Even this model overestimated gas breakthrough timing.
Following evaluation, a scheme was devised comprising a first phase of four injectors located mid-dip, to be followed by a second phase of three peripheral gas injectors after three years. But the study also revealed significant volumes of unswept oil in the top layer, down-dip of the field. So peripheral horizontal producers were recommended.
Incremental recovery from gas injection was estimated at 2% OIIP over ten years, with an average injection rate of 70 MMcf/d, representing an increase of 25% from the remaining reserves. To cater for this increase, one of the existing gas-lift trains would be replaced by a high pressure compressor, to supply 70 MMcf/d at 220 bar for reservoir injection. Also, a further compressor would be installed to recover the reproduced gas and export it through an existing pipeline.
In parallel with this work, a 3D seismic survey had been acquired to check for the presence of minor faults that might have impaired areal sweep efficiency. Also, fracture data had been collected (using formation logs in horizontal wells) to examine natural fracture distribution in the reservoir. In addition, a second tertiary gas-flood experiment was conducted in reservoir conditions on a highly permeable, composite core more representative of the reservoir.
Production facilities were installed by May 1997. Three peripheral horizontal producers were drilled followed by two more early in 1998, one located down-dip between two of the new horizontal drains. Gas injection started end-August 1997 in the first two new injector wells, with tracers put into both wells to monitor breakthrough time and gas origin. First gas breakthrough was noted end-November that year, with a significant rise in gas-oil ratio (GOR) in January 1998 on all surrounding producers. But the breakthrough times were longer than those observed on the two pilots, presenting the first positive sign of good reservoir sweep. Then fast oil rate and water cut decrease occurred, mainly on the wells situated up-dip of the injectors. More recently acquired data improved understanding of the role of fractures. Two sets of fractures were identified:
- A NW-SE trending set, parallel to the faults, always cemented and developed mainly in the structure's faulted areas
- A NE-SW oriented set, orthogonal to the faulting trend. This is mostly open and could therefore provide a conduit to oil, water and gas flow.
Permeability
According to the authors, open fractures in ABK are encountered mainly in dolomite units which exhibit moderate porosity and brittle behavior, while, limestone units are always poorly fractured. Vertical extension of the fractures is limited, while their lateral extension and connectivity cannot even be assessed. The impact may be negligible in waterflood operations, but this is not the case in gas injection.
The study team identified a strong permeability reduction at the top of the D2 layer following vertical permeability tests on several wells and reservoir pressure readings. One of the pilot injectors was sidetracked in May 1995 and perforated only in the Arab D2d and D2e layers. This resulted in change to the gas breakthrough pattern, with new breakthrough occurring only after six months on the P1 producer (despite this well being only 250 meters distant). Breakthrough was then observed on up-dip wells up to 2 km away. However, the oil recovery increase remained slow and little reduction in water cut was recorded.
Top perforations of the other injector were shut, leaving only the bottom of the D2b3 interval open. Gas breakthrough arose after only two months, but only in the up-dip producer P4. However, the change was much less marked, although an increase in the oil rate did reveal that a new part of the reservoir had been contacted. Based on this data, the bottom of the D2b3 layer was selected as the injection point on the new injectors. Completions were, however, configured for maximum flexibility in changing the injection point at a later date.
Total's conclusion was that non-miscible gas injection could boost ABK's productive life substantially. The positive impact on oil recovery should allow production from ABK to be maintained for at least another decade, according to Buresi.
Geochemistry used to evaluate ABK output
Two of ABK's main reservoirs are produced commingled. Prior to 1995, monitoring of the production contribution from each reservoir was achieved through production logging tests (PLTs) or Selective Production Tests (SPTs). At the 1997 SPE conference, various Total authors outlined how geochemistry was subsequently adopted as an alternative measurement method.
The Upper Arab reservoirs (A, B+C and D1) are produced through gas-lifted single completions with two or three zones, which allows selective production of each reservoir unit. To maximize productivity, however, most of the wells are produced commingled with all zones open.
Total's technique is based on high resolution gas chromatograms of wellhead oil samples, small quantities of which are then diluted in pentane for analysis. The key to this method is its ability to resolve small compositional differences in genetically related oils. Using special software, pairs of peaks are selected, the height ratio for which varies substantially between oil samples from different reservoirs. In a mixed oil sample, the split is derived from the value of these peak ratios.
Oil from each of ABK's reservoirs was analyzed, and the two main reservoirs were found to have a clearly different signature. Samples from commingled wells were then analyzed, and the calculated oil splits were found to be in broad agreement with values obtained from recent production logs. The method was successfully extended to wells producing from three commingled reservoirs.
Based on the results, Total concluded that the method was the only one applicable where downhole measurements are impossible, owing to completion problems. On three wells, geochemistry provided, for the first time, a means of monitoring the respective contribution of each reservoir.
The comparison between the geochemistry data and the previously assumed split revealed that the latter was frequently incorrect. Two of the wells, for instance, exhibited lower than expected contribution from the Arab D1 reservoir, leading to investigation of a potential workover.
The cost of one measurement is far below that of a single PLT operation. This, according to the authors, allows more frequent analyses to be conducted, and closer control to be maintained on reservoir performance than had been achieved previously. They planned, therefore, to analyze all wells using the geochemistry method twice a year, with the aim of reducing the frequency of downhole measurements.
Another common problem on ABK is the communication behind the casing between the completion zones, which can lead to unreliable PLT and SPST results. For instance, in 1995, a PLT on well 12 suggested crossflow downwards into the DI reservoir. Geochemistry, which is independent of this problem, established that the reservoir was, in fact, producing normally.
