Identifying Water-Flow Problems

Abstract

Summary The most important element of conformance control in fields where water injection is used to enhance oil production is identifying the nature and source of the problem. Chemical analysis and water-flow logging can help operators isolate water-flow, invasion, and annular leaks that destroy casing integrity and decrease profitability. This article presents a case history that demonstrates use of analytical techniques to identify conformance-control problems and solutions. Conditions In a west Texas waterflood field, an injection well was being used to sweep oil toward offsetting wells (Fig. 1a). Wells 1 and 2 are producers, Well 3 is the injection well, Well 4 is the water source, and Well 5 is on an adjacent lease that belongs to another operator. The waterflood zone for Well 3 is between 4,000 and 4,900 ft and between 2,500 and 3,100 ft for Wells 1, 2, and 5. Well 3 was drilled and completed to serve as an injector; however, Well 3 showed 325 psi pressure between its intermediate and surface strings soon after completion. The pressure could be relieved by opening the surface valve but would recur within 24 hours. Two remedial cement jobs were performed on the well to halt the gas flow that was causing the pressure. About 1 month after the squeeze cement job on Well 3, a wet spot (Fig. 1b) appeared on the ground around Well 4, 150 ft from Well 3. Chemical analysis of the water showed that it had not come from Well 3 or from Well 4, the injection-water source. Salt concentration in the water had increased from 15,000 to 170,000 mg/L chlorides. Additionally, the production annulus pressure in Wells 1 and 2 had increased, and shutting off Well 3 did not affect the annulus pressure of Wells 1 or 2 or the chloride content. The produced water was routed to tanks and allowed to flow so the 350- to 400-psi shut-in casing pressures could be replaced. Log Analysis A water-flow analysis log was run on Well 3 to determine water-flow velocity and direction behind two strings of casing at Well 3. This log was an oxygen-activation device that uses (1) statistical variations in isotope decay rates, through measurement of gamma ray emission counts, and (2) log movement to determine water velocity within several inches of the wellbore. Analysis of the log showed water entering the annulus at 2,807 ft, exiting at 1,980 ft, re-entering the annulus at 980 ft, exiting below the surface casing shoe at 306 ft, and rising to the surface (Fig. 2). Water-flow analysis logs appeared to show that water was coming from Well 5 and taking a route to the surface through the Well 3 annulus and a naturally occurring or induced fracture network. The water traveled through one or more brackish sands, gaining salinity by dissolving anhydrides along the way. With a most likely flow pattern established, injection into offset wells (four) was halted, including injection from Well 5, even though it belonged to another operator. Within minutes of shut-in, most annular flow had ceased. All wells remained shut in for 3 days. To isolate the source of the flow, injection was temporarily conducted into each surrounding well, one well at a time. This pumping method isolated the problem to Well 5. A foam squeeze cement job was designed and pumped into 10 new perforations between 2,692 and 2,694 ft in Well 3, successfully shutting off injection flow from Well 5. The foam cement also effectively filled and squeezed the uncemented annulus of Well 3 from 2,700 to 1,900 ft, which should help prevent future cement-sheath leaks. The offset producing wells are all back on pump and producing at presqueeze levels, and no sign of underground water escape to the surface is evident. Well 5 is now being injected at 1,300 psi, with no influence on Wells 1 through 4. P. 699