Summary Several multirate separator tests (MRTs) have been undertaken on wells in the Veslefrikk field that are on commingled production from the Brent Group and Intra Dunlin Sand. During these tests, produced-water samples were also collected. Integrated analysis of the results of interpretion of the produced-water analyses and the MRT results have provided a range of information for each production zone including the nature and composition of the produced water, seawater fraction of these produced waters, fraction of total water flow being produced, pressure, productivity index, oil and water rates, and water cut. This information can reduce the need for deploying production-logging tools (PLTs), allows the scaling potential between the deeper and the shallower zones to be evaluated, aids squeeze-treatment design, is beneficial for predicting formation damage from crossflow, and aids water-shutoff decisions. In this paper, the methods of interpreting the produced-water analyses are presented through the use of a field example. To aid interpretation, interpretation of other produced-water analyses from the field and reactive-transport modeling have been undertaken to better understand reactions occurring in the reservoir as a result of seawater injection and the effects of these reactions on produced-water compositions. In an accompanying paper (Tjomsland et al. 2011), the integrated MRT-analysis techniques are described and the results and applications of additional field examples are presented.

Produced water analyses from the Birch Field, UK North Sea have been interpreted and combined with simulation results to explain the causes of changes in produced water compositions over time. Preliminary conclusions are that two formation waters are present in the oil leg, both trapped at the time of oil emplacement. Lower salinity formation water has been expelled from the Kimmeridge Clay Formation (KCF) and dominates shallower sections of the reservoir. Higher salinity formation water is thought to be ancient Brae aquifer water and dominates the deeper sections. Some lateral variability in formation water compositions is evident. Produced water from individual wells is a mixture of the lower salinity formation water, higher salinity formation water and injected water. Trends in produced water compositions over time reflect a relative decrease in formation water production and increase in injection water production. Depending on the constituent, reactions occurring as a result of injection of water into the reservoir also affect the composition of produced water. The results have challenged previous concepts relating to water production at Birch and will be considered in scale management plans in future. They can also be used to constrain reservoir simulations, to aid enhanced oil recovery decisions and to provide more reliable tracking of injection water and formation waters entering the production wells at Birch.More generally, this study has demonstrated the importance of evaluating produced water analyses as early as possible after water breakthrough. Integration of reservoir simulation studies with the interpretation of produced water analyses can provide information that benefits scale management, STOIIP calculations, reservoir models, and tracking of injection water as well as providing analogue information that can help reduce uncertainties associated with the development of deep water and marginal subsea fields. TX 75083-3836, U.S.A., fax 01-972-952-9435.

The Blane Field, North Sea, has one injection well and two production wells and is tied back to the Ula Field platform. The original scaling risk assessment was based on injection of platform produced water (PW) with minor seawater (SW) (~90:10 PW:SW). However, after injection of 25:75 PW:SW for only 6 weeks, a change in operational circumstances on the Ula Field meant that only 10:90 PW:SW injection water could be supplied for the next 18 months. There was a risk that this might result in unmanageable BaSO4 scaling conditions in the production wells but the alternative would be to cease injection, leading to reservoir pressure decline and loss of oil revenues. The need for a rapid decision negated the use of reactive transport reservoir simulations to predict the future BaSO4 scaling risk under the new injection scenario so a novel, alternative approach was adopted. A history matched ECLIPSE model served as the basis for predicting the types of water entering the production wells over time and their rates. A 1-D reactive transport model was then used to predict the Cl, Ba and SO4 composition of these waters after accounting for the effects of reservoir reactions. These results were integrated in a spreadsheet to provide predictions of Cl, Ba and SO4 concentrations in the produced water from each well over time. The results for future injection water scenarios indicated that the scaling risk would increase over time in the wells but, due to deposition of BaSO4 and CaSO4 in the reservoir, the BaSO4 scaling risk would be manageable even allowing for uncertainties associated with this approach. Based on these results, and those of associated studies, a decision was made to continue water injection resulting in avoidance of loss in oil revenues. This novel scaling prediction approach may be useful on other fields where reactive transport reservoir simulations may not be possible.

Produced water analyses from the Birch Field, UK North Sea have been interpreted and combined with simulation results to explain the causes of changes in produced water compositions over time. Preliminary conclusions are that two formation waters are present in the oil leg, both trapped at the time of oil emplacement. Lower salinity formation water has been expelled from the Kimmeridge Clay Formation (KCF) and dominates shallower sections of the reservoir. Higher salinity formation water is thought to be ancient Brae aquifer water and dominates the deeper sections. Some lateral variability in formation water compositions is evident. Produced water from individual wells is a mixture of the lower salinity formation water, higher salinity formation water and injected water. Trends in produced water compositions over time reflect a relative decrease in formation water production and increase in injection water production. Depending on the constituent, reactions occurring as a result of injection of water into the reservoir also affect the composition of produced water. The results have challenged previous concepts relating to water production at Birch and will be considered in scale management plans in future. They can also be used to constrain reservoir simulations, to aid enhanced oil recovery decisions and to provide more reliable tracking of injection water and formation waters entering the production wells at Birch.More generally, this study has demonstrated the importance of evaluating produced water analyses as early as possible after water breakthrough. Integration of reservoir simulation studies with the interpretation of produced water analyses can provide information that benefits scale management, STOIIP calculations, reservoir models, and tracking of injection water as well as providing analogue information that can help reduce uncertainties associated with the development of deep water and marginal subsea fields. TX 75083-3836, U.S.A., fax 01-972-952-9435.

No abstract