The success of many oilfields depends upon efficient disposal of produced water. In the case of the Masila block in Yemen, the difficulty of water disposal was underestimated because the high permeability of the production wells led to expectations that high injection rates and economic subsurface disposal would be easily attained.

A number of standard approaches were taken to obtain injection rates of 40,000 BWPD per well. After initial failures, a multi-disciplined team was formed to review the practical and theoretical aspects of formation damage and water quality. Successful injection at rates exceeding 80,000 BWPD per well were subsequently achieved by underbalanced drilling and improving water quality. Two orders of magnitude of increase in injectivity index. Optimizing the drilling and completion methods saved $14MMUS that would have been needed to dispose of water at hydraulic fracturing pressures.

The learning curve and dead ends that led to the successful subsurface disposal of over 200,000 BWPD are described. Perforating, core flow tests, core analysis, corrosion, sand behavior, pressure measurement, drilling fluids, completion configuration, filtration and remedial stimulation are discussed.


Exploration rights to Yemen's Masila block were granted in 1987. The original block covered 8.9 million acres, the significant portion of which is shown in Figure 1. The initial discovery wells were drilled early in 1991, and commerciality was declared in December, 1991 from the Camaal, Heijah and Sunah Fields. In 1992, the additional fields of Hemiar, and Tawila were discovered and oil production started in July, 1993.

The main oil producing reservoir is the Qishn formation, and typical reservoir characteristics are shown in Table 1. Water production was anticipated early in the producing life of the Qishn sands, and due to the good permeability of these sands, few problems were anticipated in re-injecting produced water in the Qishn. The water production forecast in late 1993 is shown in Figure 2. Water production was (and still is) expected to peak in 1997–1998, at 350,000 – 400,000 BWPD. Assuming disposal in the Qishn, and an injection tubing head pressure of 100 psi, Figure 3 shows the effect of different sizes of injection tubulars as a function of Injectivity Index (II). Table 2 shows the injector well requirements, as a function of II. Clearly, sustained II's of 100–300 were necessary to minimize the number of disposal wells. In the lower part of the Qishn, the S3 typically has permeability of up to 5 darcies, and for this reason 2 wells were initially selected for produced water disposal into this interval, CPF SWD#1 and Camaal SWD#1. Figure 4 and Figure 5 show the initial completion diagrams for these wells.

Initial Problems With Water Injectors

CPF SWD#1. CPF SWD#1 was perforated in August, 1993 with tubing conveyed perforating equipment, using 500 psi underbalance, 4 1/2" TCP guns, 5 SPF, 37 g. charges in a filtered 3% KCl solution of produced Qishn water. Typical Qishn water composition is given in Table 3. It can be seen that the Qishn water is fresh. After perforating the S3 sand, an electrical submersible pump (ESP) was run to pump the well and obtain formation water for re-injection. Figure 4 shows the ESP and pressure gauge arrangement. The well was pumped at 12,300 BWPD for a total of 5 hours, resulting in a final pressure drawdown of 27 psi. This indicated a productivity index (PI) of 456 BWPDI psi.. The build-up data were analyzed to show no near wellbore damage and no indication of a limited reservoir.

This content is only available via PDF.
You can access this article if you purchase or spend a download.