This paper was prepared for presentation at the 1999 SPE European Formation Damage Conference held in The Hague, The Netherlands, 31 May–1 June 1999.

Water is very often associated to the oil production, for geological reasons, but also because it is the most frequent mean of secondary recovery. However all fields are not comparable in their behaviour. In the best cases water is effectively contributing to the oil sweeping and the bulk of the oil reserves can be produced at low water cuts. In other cases, it can be said that water is inevitable to the oil and huge volumes of high water liquids must be lifted from an early stage to produce the oil. In some cases, mechanical or chemical water shut off techniques can help to reduce the water production however, depending on the specific conditions, they are not always cost effective, their implementation can be tricky, and their efficiency may be limited in time. Therefore the operator is often left with the standard solution of upgrading its field and process facilities to cope with the produced water constraints. However, due to the increase difficulties resulting from the drastic new environmental regulations, the operators tend to focus more than before on the produced water associated cost. A typical offshore field production history and capex and opex breakdown was analysed to highlight the impact of the water on a field economy. The new interest linked to the emerging technology of downhole separation and re-injection has motivated the testing of a DOWS unit on the well LA-90 in the Lacq Superieur field in France and results of this operation are presented. Considering the shortcomings in the existing static cyclone technology which is implemented in the down hole separation systems, TFE has undertaken since three years a R&D program based on an innovative concept of rotary cyclone. The base of the theory and its implementation are presented along with the promising preliminary results. Introduction In many places, water is inevitable for the oil. Although initial oil production from a field is often dry, the water often invites itself in the well at some stage, sometimes much earlier than initialy anticipated. Some reservoirs are connected to large aquifers providing a strong pressure support to the oil production. Depending on the geology of the structure and on the reservoir characteristics, different schemes can account for the water production process. In the bottom drive reservoirs, where the water is directly underlaying the oil, the water coning, resulting from the pressure drawdown applied to the formation, is governing the water production. In this type of configuration, the critical rate per well is generally too low to be economical. In fact, there are little reservoirs where an efficient gravity drainage can be implemented. For the edge drive reservoirs, production wells are drilled much far away from the oil and water contact, but water tends to channel faster through high permeability drains and reaches the producers sometimes very early in the life of the field. When the oil layer is only connected to a small aquifer, there is not a sufficient pressure support to compensate for the oil production. Then the reservoir pressure is decreasing with time, which is often very detrimental to the ultimate oil recovery. Hence, a pressure maintenance scheme is required and water (or gaz) has to be injected into the reservoir to balance the oil offtake. Depending again on the reservoir characteristics and on the geology, the water injection wells can be located at the periphery of the oil layer, away from the producers, or on the contrary they must be drilled between the oil wells. Obviously, the same comments as before, concerning the risk of early water production, can apply to these schemes. In all cases, an early water breaktrough results in a reduce sweeping efficiency which has a negative impact on the oil recovery. This is even aggravated by the unfavourable mobility ratio between the oil and the water, since the viscosity is generally higher for the oil than for the water. Therefore, a longer production period is required in order to make up for the delayed oil and huge cumulative volumes of water are produced ultimately.

Produced water re-injection (PWRI) for use in pressure maintenance is an alternative to discharge to the environment and is increasingly practised industrially in view of the combination of ever more stringent discharge standards and high water treatment costs. Due to the damaging effects on formations of produced waters it is often necessary to inject in fracturing regime to maintain the desired injection rate. Then the main issues of the process are the impact on well injectivities, fracture growth and sweep efficiency. All these issues are addressed by presenting an onshore field case of PWRI on a low permeability carbonaceous reservoir. Aquifer water has been injected for ten years and progressively substituted by produced waters for the next ten years. This case is well monitored and documented (fall-off and step rate tests, temperature logs, water quality). The main parameters affecting the process are illustrated in diagnostic analysis and the injectivities are satisfactorily reproduced using different simulation techniques. Special care is taken to describe filter cakes properties and distributions behind the fracture faces and inside the de fracture in order to well define the effects on the loss of fracture injectivity. The resulting fracture dimensions are very consistent with the last fall-off measurements and analysis. Finally simulating incrementally this fracture growth in the Eclipse reservoir model it was possible to demonstrate that water breakthrough occurs when the fracture tip reaches laterally higher permeability zones in the producer area and to correlate decreases of water production with injection shut in periods associated with fracture tip closure. Introduction The case presented is an onshore field at 3000 m depth and situated in the south west of France. The reservoir is a very hard carbonate with a Young modulus of 500000 bars. Very significant lateral heterogeneities are existing. The permeabilities in the injection area are less than 1 mD compare to 1–10 mD in the producing area with effective permeabilities of 50–2100 mD due to a natural fissure and fracture network. The basic data at the beginning of injection on 1979 for one typical well are given in table 1. Fresh aquifer waters from another field has been injected peripherial (figure1a) for pressure support at high pressures during 10 years followed by commingle produced water reinjection during another 10 years. Due to the strange behavior of the well different monitoring techniques has been used over the life of the field leading to interpretation difficulties. With the help of the recent improved knowledge on fractured injection regimes and reinjection effects (Ref. 1 to 7) all the field data has been reprocessed and the results of these analysis are discussed in the following chapters. Historical diagnostic analysis The entire historical is presented in figure 1.The different injection periods and tests are mentioned. The first surprise is that at first look there is apparently no effect of produced water re-injection on injectivity. Transforming surface in bottom-hole pressure does not change this observation. Reservoir pressure effects. In fact the reservoir pressure is decreasing significantly during the produced water injection period as shown on figure 2. Then plotting the evolution of the differential pressure (bottom-hole pressure - reservoir pressure) with time on figure 3 shows a significant decrease in injectivity during this last period (increase in pressure for the same rate of 400 m3/d). Pressure rate plot. When the reservoir pressure is varying this plot must also be done using differential pressures otherwise this can lead to big mistakes in the diagnostic of injection regimes. Figure 4 shows that the injection regime is fracturing excepted at the early beginning of injection where the points close to the Y axe correspond to radial flow. Reservoir pressure effects. In fact the reservoir pressure is decreasing significantly during the produced water injection period as shown on figure 2. Then plotting the evolution of the differential pressure (bottom-hole pressure - reservoir pressure) with time on figure 3 shows a significant decrease in injectivity during this last period (increase in pressure for the same rate of 400 m3/d). Pressure rate plot. When the reservoir pressure is varying this plot must also be done using differential pressures otherwise this can lead to big mistakes in the diagnostic of injection regimes. Figure 4 shows that the injection regime is fracturing excepted at the early beginning of injection where the points close to the Y axe correspond to radial flow.

This paper was prepared for presentation at the 1999 SPE European Formation Damage Conference held in The Hague, The Netherlands, 31 May–1 June 1999.

No abstract