Abundant water chemistry analyses from nine different locations (predominantly petroleum reservoirs) on five continents were evaluated. This information, together with local mineralogy, depth and temperature relations provided a sound basis from which to investigate the most important controls on formation water composition. In particular, the detailed study of two very different hydrocarbon reservoir case studies (the Central US coalbed methane reservoir, the San Juan Basin and the North Sea oilfield Miller) provided an insight not only into the fundamental controls on formation water composition, but also into the effects of active oilfield development on systems that are very sensitive to change on rapid timescales.
The geochemistry of San Juan waters is controlled by the introduction of bicarbonate through carbonate dissolution and methane/coal oxidation leading to leaching of Na-bearing clay minerals, and by ion exchange on clay minerals and dilution by meteoric waters in certain locations. The time series of produced waters from Miller enabled detailed study of fluid mixing in the field and the physical, chemical and thermodynamic response of the system to the
injection of seawater. Changes occur in the concentrations of many water components through time that cannot be explained by linear mixing between formation water and injected water and require dissolution or precipitation reactions to have occurred between injection and production
sites. For example Ba, and SO4 concentrations are affected by equilibrium with barite and what is likely to be sulphate reduction. Also, excess Si present in the fluid is due to dissolution of the silicate phases in the reservoir, and demonstrates reactions between silicate minerals occur on a fast enough timescale to buffer the pH of the water.
Integration of all available data shows consistent patterns of behaviour, which implicate mineral-fluid interactions in the subsurface as a major control on formation water chemistry. For example, globally, Ca concentrations are shown to behave in one of three ways, all of which depend on water interaction with the host rock, be it silicate or carbonate, clastic or evaporite. Distinct trends arise for bicarbonate waters, brines derived by halite dissolution and formation brines that have evolved extensively with silicates. In addition, K concentrations are closely
related to feldspar-clay equilibria and Mg concentrations are influenced predominantly by carbonate minerals with significant contribution from clays. It is likely that initial Ba concentration is related to interaction with K-feldspars and SO4 is controlled by equilibrium with sulphate mineral phases as well as by redox.
A greater understanding of formation water chemistry leads to an improved perception of the importance of these systems in terms of both furthering scientific progress and the technological development of the oil and gas industry. In particular, produced water chemistry analyses from
Miller were used to appraise and improve the most important aspects of both generic and specific reservoir models. A set of simple models emphasised the point that small variations in reservoir property parameters can have significant effects on model outputs, and thus the
highlighted the importance of thorough reservoir characterisation, particularly permeability heterogeneity, capillary pressure and relative fluid permeabilities.
Geochemical models of three different systems from the integrated database (the Alberta Basin, a Colombian onshore oilfield and an oilfield from offshore Gulf of Mexico) illustrate that reservoir rocks containing a wide variety of minerals are the most effective at limiting pH decrease following the injection of CO2 into the system. The geochemistry, in particular the salinity, of the formation water present also has a significant bearing on the processes that are likely to occur during CO2 sequestration.