Abstract Bacterial contamination of injection waters is described as only one of many complex water treating problems. The need for test data other than bacterial counts for establishing bactericide treating requirements is stressed. Some of the adverse operating conditions under which bactericides must function in order to be effective are presented. Case histories are shown for four waterflood projects. Detailed test data for three of them are included in graphical form. Possibilities for reducing bacterial contamination by mechanical means such as separating commingled waters are discussed. Economics of treatment used on these floods are included. Introduction Waterflood bacteriology has progressed from an interesting curiosity to a specialized science within the last 12 years. Extensive effort by many companies, a few committees and many individuals has provided detailed standardized methods for investigating bacterial activity. Through the years, the bacteriologist refined his techniques, gave us new bactericides and brought the solutions to some of our problems into sharper focus. Now he is examining an ever increasing number of water-treatment problems to determine whether they are related to or caused by bacterial activity. This is as it should be. The base of knowledge must continue to be enlarged. However, study in this direction may tend to overemphasize the role of bacteria in water treatment problems. Those responsible for water treatment should be cognizant of the fact that elimination or reduction of bacterial activity in waterflooding is justified only if it is the most economical method of controlling corrosion and plugging. Therefore, they must be able to detect and determine the cause o corrosion or plugging tendencies. We do not believe that this is possible by investigating bacterial activity alone. Many other tests are required, no one of which is universally more dependable than the others. An entire group or package of tests results is needed. For the purpose of this discussion, results of several test are shown for three floods, labeled as Floods A, B and C, and results of coupon corrosion tests only are shown for a fourth project, Flood D. All are closed systems. In comparing these test results it will be noted that a trend or change often is distinctly detectable by results of one or more series of tests but is completely obscured in results of other tests. Injection water in Flood A is a mixture of produced brine and fresh make-up water. Bacterial growth was high enough to cause significant corrosion. The flood was operated three years without appreciable chemical treatment. Adequate test data are available for this period. Injection water in Flood B is a mixture of produced brine and fresh make-up water. Bacterial contamination was serious, and continuous bactericide treatment was used for three years. In the fall of 1961, treatment was changed to daily batches of corrosion inhibitor with bactericide batch treatment about once every three months. Make-up water in Flood C is fresh water from river-bank wells. After one year, the produced salt water and the fresh water were separated and bactericide was added to the produced water only. Valve leakage resulted in inadvertent mixing of the fresh and salt waters for several months in 1962, causing a sharp increase of the corrosion rate. Make-up water in Flood D is a deep-well brine. Test data are available for the last six years of the 1 1 years of operation. Moderate sulfate reducer counts were found with sulfide contents ranging from zero to 0.3 mg/liter. Satisfactory control of internal corrosion was obtained by corrosion inhibitor treatment without a bactericide. These four floods were selected for discussion because of unique difficulties encountered. They are not "text book" cases of bacterial problems which are easily evaluated and solved. However, they are representative examples of the following general statements:bacterial tests alone do not provide sufficient data to determine the need for treatment,concentrations of bacteria in the flowing water often show poor correlation with other test results,treatment with a bactericide is not always required or economically feasible, eve though bacterial tests indicate considerable contamination,bactericides which are effective only under ideal operating conditions will not provide adequate bacteria control in most water floods,upper limitations on bactericide concentrations usually are controlled by treatment cost rather than kill requirements, andbacterial growths usually are more severe in systems with commingled fresh and salt waters than in systems where the fresh makeup water and the produced brine are handled in separate systems. JPT P. 1073^