Disclaimer for a Provisional Database: "The data you have secured from the USGS database identified as have not received Director's approval and as such are provisional and subject to revision. The data are released on the condition that neither the USGS nor the United States Government may be held liable for any damages resulting from its authorized or unauthorized use." Limitations: The information in the produced water database should be used with careful consideration of its limitations. The database is considered sufficiently accurate to provide a indication of tendencies in water composition from geographically and geologically defined areas. It is not appropriate for depiction of modern produced water compositions or examination of trends on small scales. The USGS makes not warranty regarding the accuracy or completeness of information presented in this database. Specific aspects of the database that should be considered are listed below. * Much of the information in the database cannot be independently verified. Methods of collection, sample preservation, analysis, assignment of geologic units and record keeping were not rigorous or standardized. Because of these uncertainties, users are advised to check data for inconsistencies, outliers and obviously flawed information. * Methods of well construction, sample collection and chemical analysis have changed over time. * The distribution and relative amount of water produced within a province and among geologic units may not be represented by the samples in the database. No sampling was planned to accurately depict the aggregate water composition of any area whether it be province, state, county or field. * The geologic units developed for petroleum production have changed over time. Data from a province collected 30 years ago may not resemble current production. * The composition of produced water within a province, field or even well may change in time as a result of water flooding, recompletion in other intervals, and workovers. * Water samples are commonly collected when a well has production problems or during the initial development of a well. Although criteria were applied to remove the obviously contaminated samples, the culling of unrepresentative data is considered incomplete. * Although most obvious redundant entries were removed from the database, many of the records represent multiple samples of the same well. Therefore aggregate statistics may be weighted by relatively few wells.