Waterflooding

Content on this page requires a newer version of Adobe Flash Player.

Get Adobe Flash player

Waterflood Design and Implementation

Waterflooding is the most successful and widely used enhanced oil recovery process. This is because water is widely available and inexpensive relative to other fluids, easy to inject, and highly efficient in displacing oil.

A key factor in the design of a waterflood is estimation of the oil recovery. This in turn is dependent on reservoir characteristics and the manner of project implementation and operation.

Waterflood Oil Recovery


Waterflood recovery, NP, can be computed at any time in the life of a project by using the formula:

Waterflooding

where
N = the oil in place in the floodable pore volume at start of water injection.
ED = displacement efficiency (fraction of oil saturation at the start of water injection which is displaced by water in the invaded zone)
EA = areal sweep efficiency (fraction of floodable pore volume area swept by the injected water)
EV = vertical sweep efficiency (fraction of the floodable pore volume in the vertical plane swept by the injected water)

Note: Floodable pore volume is highly dependent on net pay discriminators such as permeability and porosity cutoffs.

Reservoir properties that significantly affect these efficiency factors are:

    • Mobile oil saturation,
    • Mobility ratio, and
    • Reservoir heterogeneity

Mobile Oil Saturation

Mobile oil saturation is the oil saturation at the start of the waterflood, So, minus the residual oil saturation to water, Sorw.
Higher values of mobile oil saturation results in higher waterflood oil recovery.

Mobility Ratio, M

Mobility ratio, M, is the water mobility in the water invaded portion of the reservoir divided by the oil mobility in the non-contacted portion. In turn, the mobility of the fluid is the permeability of the rock to that fluid divided by the fluid viscosity. Hence, in terms of relative permeability:

Waterflooding

where
    krw = water relative permeability at the average water saturation in the swept zone
    µw = water viscosity
    kro = oil relative permeability in the oil bank ahead of the water (normally equivalent to 1.0)
    µo = oil viscosity


Waterflooding
Figure W-1. Effect of oil viscosity and rock wettability on mobility ratio, water viscosity = 0.5 cp.
(Interstate Oil Compact Commission, 1983).


It is important to mention that the manner in which water displaces oil from reservoir rock, i.e. relative permeability effects, depends upon the preferentially wettability of that rock. Figure W-1 shows water-oil mobility ratio as a function of oil viscosity and rock wettability. Mobility ratio increases with oil viscosity and is higher for oil-wet rocks.

Displacement, areal and vertical sweep efficiencies all increase as mobility ratio decreases.

Reservoir Heterogeneity, V

Dykstra and Parsons (1950) introduced the concept of the coefficient permeability variation V, which is designed to describe the degree of heterogeneity within the reservoir. The value of this uniformity coefficient ranges between 0.0 for a completely homogeneous system and 1.0 for a completely heterogeneous system.
Areal or vertical permeability variation is computed by first arranging all relevant core permeability values from maximum to minimum, and plotting each versus the percentage of the total number of samples having higher permeability. The plot is made on log-probability coordinates and a straight-line is expected if the permeability data is log-normally distributed. Coefficient of permeability variation, V is computed as follows:

Waterflooding

where,

k50 = permeability at the norm or 50% probability
k84.1 = permeability (one statistical deviation away from the norm) at 84.1% probability

Vertical sweep efficiency increases as vertical reservoir heterogeneity (stratification) decreases.


Displacement Efficient


The oil displacement efficiency of a waterflood is determined by water-oil relative permeability characteristics, oil and water viscosities, capillary pressure, and reservoir dip.

One established procedure for estimating displacement efficiency is to construct a plot of the fractional flow of water versus water saturation. Consider the general case of displacement of oil by water in a one-dimensional system inclined at an angle α to the horizontal as shown in Figure W-2.

Figure W-2 Waterflooding
Figure W-2. One dimensional displacement.

The analytical procedure (Buckley-Leverette, 1942) assumes that immiscible displacement can be modelled based on the idea of the so-called “leaky piston”. This means that there is a considerable amount of by-passing of oil by the flood front. Equation W.4 (in the Darcy system of units) gives the fractional flow of water, fw, at any position in the linear system.



Wateflooding
Where
               k = permeability
               kro = relative permeability to oil
               krw = relative permeability to water
               µo = oil viscosity
               µw = water viscosity
               q = total flow rate
               Pcow = oil-water capillary pressure
               Δρ = water-oil density difference
               α = angle of reservoir inclination/dip to the horizontal
               x = distance along direction of displacement

 

Waterflooding

Since oil and water relative permeabilities vary with water saturation, the fractional flow of water is related to water saturation.

Figure W-3 is a plot of fw as a function of water saturation. The oil displacement efficiency at water breakthrough is found from a tangent to the fractional flow curve drawn from the point corresponding to fw = 0.0 and the connate water saturation Swc. Three important features of the displacement process are derived from this tangent:

  • The water saturation at the point of tangency, Swf, is that at the flood front and, consequently, that at the producing well at the time of breakthrough.
  • The producing water-cut at breakthrough is given by the value of the fractional flow at the point of tangency, i.e. fwf.
  • The average water saturation in the water invaded portion of the reservoir at the time of breakthrough is the value of the water saturation where the tangent intersects fw = 1.0, i.e. Swbt.
The oil displacement efficiency at breakthrough is calculated from

Waterflooding

Figures W-4 and W-5 show the effects of oil viscosity and rock wettability on waterflood oil displacement efficiency. It is apparent that for a given reservoir rock the breakthrough displacement efficiency is higher the lower the oil viscosity.

For strongly water-wet rock, the tangent to the fractional flow curve at oil viscosities of 2.0 cp and lower passes through fw = 1.0 at the maximum water saturation. Under these conditions there is “piston-like” displacement of the oil by water; that is, all the recoverable oil in the swept portion of the reservoir is displaced at breakthrough.

When the displacement is not piston-like, the average water saturation in the invaded portion of the rock increases with continued injection, and the displacement efficiency correspondingly increases. Note that for higher oil viscosities, the amount of oil recovered after breakthrough increases. Also, the amount of oil recovered after breakthrough is larger in a strongly oil-wet reservoir than in a strongly water-wet reservoir containing the same oil viscosity.

Figures W-4 & 5 Waterflooding

The maximum oil displacement efficiency, EDmax, is given by

Waterflooding

Areal Sweep Efficiency, EA


Figure W-6 Waterflooding

Figure W-6. Areal sweep efficiency as a function of mobility ratio (M) and producing water-cut (fD), for a five-spot pattern (Caudle and White, 1959).

Areal sweep efficiency of a waterflood depends upon two main factors: the mobility ratio and the flooding pattern or well arrangement.

Figure W-6 shows areal sweep efficiency for a five-spot pattern as a function of mobility ratio and producing water-cut (fD). For any given producing water-cut, lower mobility ratios result in higher areal sweep.

Additionally, for any given mobility ratio, areal sweep increases after breakthrough (i.e. fD > 0.0) as producing water-cut increases. This increase in areal sweep after breakthrough diminishes considerably at high values of mobility ratio, approaching 1000. At these high mobility ratios, areal sweep converge asymptotically around 40%.

In thin formations it may be feasible to waterflood heavier oils with parallel horizontal wells. The lower displacement efficiency will be offset by a much higher areal sweep. Here vertical sweep will be less of an issue.

Vertical Sweep Efficiency, EA


Figure W-7 Waterflooding Figure W-7. Vertical sweep efficiency as a function of permeability variation, V, and mobility ratio, M. Producing water-cut = 98%. (Dykstra and Parson, 1950).

Mobility ratio, an important influence on areal sweep of a waterflood, also plays a large part in the vertical sweep.

Figure W-7 shows vertical sweep efficiency as a function of vertical permeability variation coefficient, V, and mobility ratio, M. As the reservoir becomes more vertically heterogeneous or stratified, vertical sweep is expected to decrease.

For any given degree of heterogeneity, vertical sweep decreases as mobility ratio increases. Note: Mobility ratio has a diminishing effect on vertical sweep, as the formation becomes more homogeneous vertically or less stratified.

Factors to Consider in Waterflooding

Reservoir Geometry

The areal geometry of the reservoir will influence the location of wells and, if offshore, the location and number of platforms required. Geometry will also dictate the methods by which a reservoir can be produced through water injection practices.

Analysis of reservoir geometry and past production performance is often important when defining the presence and strength of a natural water-drive and the need to supplement the natural water influx.

Fluid Saturations

In determining the suitability of a reservoir for waterflooding, a high mobile oil saturation is the primary criterion for successful flooding operations.


Fluid Properties

The viscosity of the crude oil is considered the most important fluid property that affects the degree of success of a waterflood project. It has the import effect of determining the mobility ratio that, in turn, controls sweep efficiency.


Reservoir Depth

Reservoir depth affects both technical and economic aspects of waterflooding.
Maximum injection pressure will increase with depth. On the other hand, a shallow reservoir imposes a constraint on the injection pressure that can be used, because this must be less than fracture pressure. An operational pressure gradient of 0.7 psi/ft of depth will provide a sufficient margin of safety.
The cost of lifting oil from very deep wells will limit the maximum economic water-oil ratios that can be tolerated, thereby reducing the ultimate recovery factor and increasing the total project cost per barrel.


Lithology and Rock Properties

Lithology has a profound influence on the efficiency of water injection. In some complex reservoirs, only a small portion of the total porosity, such as fracture porosity, will have sufficient permeability to be effective in water injection operations. In these cases, a water injection program will have only a minor impact on the matrix porosity, which might be crystalline, granular, or vugular in nature.
Clay minerals in some sands may clog the pores by swelling and deflocculating when waterflooding is used.


Reservoir Uniformity and Pay Continuity

Areal continuity of the pay zone is a prerequisite for a successful waterflooding project. Isolated lenses may be effectively depleted by a single well completion, but any flooding process requires that both the injector and producer be present in the lens. Breaks in pay continuity and reservoir anisotropy caused by depositional conditions, fractures, or faulting need to be identified before determining the proper well spacing and flood pattern orientation.
If the formation contains a stratum of limited thickness and with very high permeability (i.e. a thief zone), rapid channelling and bypassing will develop. Unless this zone can be located and shut off, producing water-oil ratios will soon become too high for the flooding operation to be considered profitable.


Primary Reservoir Drive Mechanism

Primary drive mechanism and anticipated ultimate oil recovery should be considered when reviewing possible waterflood prospects.

Water-drive reservoirs that are classified as active water-drive are not usually considered to be good candidates for waterflooding because of the natural ongoing water influx. However, in some instances a natural water drive could be supplemented by water injection in order to:

  • Support a higher withdrawal rate or better balance voidage and influx volumes.
  • Better distribute the water volume to different areas of the field to achieve more uniform coverage.

Gas-cap reservoirs are not normally good waterflood prospects because the primary mechanism may be quite efficient without water injection. However, the presence of a gas cap does not always mean that an effective gas cap is functioning. If the vertical communication between the gas cap and the oil zone due to low vertical permeability, a waterflood may be appropriate in this case.

Smaller gas-cap drives may be considered for water injection, but the existence of the gas cap will require greater care to prevent migration of displaces oil into the gas cap. This migration would result in loss of recoverable oil due to the establishment of residual oil saturation in pore volume which previously had none. If the gas cap is repressured with water, a substantial volume may be required for this purpose, thereby lengthening the project life and requiring a higher volume of water.


Solution gas-drive reservoirs generally are considered the best candidates for waterfloods. Because primary recovery is low, the potential exists for substantial additional recovery by water injection. As a general guideline, waterfloods in solution gas-drive reservoirs frequently will recover an additional amount of oil equal to primary recovery.

Volumetric undersaturated reservoirs depend on rock and liquid as the main driving mechanism. In most cases, this mechanism will not recover more than 5% of the original oil in place. These reservoirs offer an opportunity for greatly increasing recoverable reservoirs.

Waterflood Implementation


Probability of economic success is improved if the following operational guidelines are adopted.
 

    • Start the waterflood early in the field's life
    • Understand the reservoir geology
    • Infill drill to reduce lateral pay discontinuity
    • Develop the field with a waterflood pattern that uses geological architecture  to the best advantage
    • Have all of the pay open in both injection wells and producing wells
    • Keep all producing wells pumped off
    • Inject below formation fracture pressure
    • Inject clean water
    • Operate the waterflood based on injection well tests
    • Conduct a surveillance program

Start the waterflood early in the field's life


Reservoir pressure drop below bubble-point pressure during primary depletion causes gas to come out of solution in the reservoir. During waterflooding, any free gas saturation must be collapsed before waterflood response can occur. This delay in waterflood response can hurt the economics of the project. The loss of solution gas from the oil also increases oil viscosity, which lowers oil flow rate and adversely affects mobility ratio. If the reservoir reaches as advance state of depletion prior to the start of the water injection, the project may fail.

It should be noted, however, that in water-wet reservoirs the waterflood recovery may be improved if some (optimum) gas saturation is present at the start of the flood (Khelil, 1967).

Understand the reservoir geology


It is critical that the reservoir geology be well understood. A good suite of openhole logs, including gamma ray, neutron porosity, density porosity, acoustic porosity (for synthetic tie back to 2-D and 3-D seismic data), deep and micro resistivity logs are a must.

Besides logs, a good areal distribution of whole cores that cut the entire section are necessary to define the permeability distributions throughout the field. If the reservoir is believed to be fractured, an oriented core can reveal the orientation of the fractures. This knowledge will be useful for orientation of the waterflood patterns.

Carefully controlled drawdown tests followed by buildup tests conducted early in the life of the field are useful for determining reservoir characteristics such as permeability-thickness, drainage radius, distance to reservoir boundaries, and existence of dual porosity. These tests should be conducted before any free gas saturation is established and in as many wells as possible.

Stratification in reservoirs can create major problems, leading to low vertical sweep efficiencies. Thin, high-permeability channels in stratified reservoirs prevent efficient flooding of other zones.

Infill drill to reduce lateral pay discontinuity

Lateral discontinuities in reservoir pay will cause oil to be inaccessible to the sweeping action of injected water. Reducing inter-well distance through infill drilling will improve reservoir continuity, areal and vertical sweep efficiencies, and therefore oil recovery. Note: if there is no increase in reservoir continuity the producing rate will increase but ultimate oil recovery would remain unchanged.

Starting a waterflood at one well density and then down-spacing to a higher well density can have adverse consequences. The higher water mobility from the original injection patterns may lead to water cycling on one side of a pattern, while the other side is in the prime of its waterflood response. Therefore, infill drilling prior to starting a waterflood advisable.

Increased environmental concerns about injection wellbore integrity (i.e. corroded casing that may not have cement behind the pipe) and the cost of moving pumping units and their electrical service support drilling new injection wells instead of converting old producers, whenever possible.

Develop the field with a waterflood pattern that uses geological architecture to the best advantage

The waterflood or pattern layout should be selected is such a manner that

  • maximizes oil recovery,
  • minimizes water production, and
  • minimizes the number of new wells required.

Peripheral water injection is appropriate where the formation permeability is large enough to permit the movement of the injected water at the desired rate over the distance of several well spacings from injection wells to the last line of producers.

In basal injection, the fluid is injected at the bottom of the structure. Many water-injection projects use basal injection wells with additional benefits being gained from gravity segregation.

Heterogeneous reservoirs will not respond well to non-pattern waterflooding. The low injectivity associated with heterogeneous formations will not allow injection to keep up with fluid withdrawals, let alone pressurize the reservoir to drive any free gas back into solution. Patterns should have a producer-to-injector ratio of 1:1. Inverted 9-spot patterns with their 3:1 producer-to-injector ratio have not worked as well as 5-spot and line-drive patterns. Orientation (aspect ratio) of the patterns can be critical if there is a preferential permeability direction or natural fracturing.

Have all of the pay initially open in both injection wells and producing wells


Both production and injection wells should be completed across all of the hydrocarbon bearing rock, not just the highest porosity streaks, otherwise oil will be bypassed. Note that once injection has been established in a wellbore, it may be difficult to effectively open and stimulate new intervals due to relative permeability effects: aqueous fluids tend to enter high water saturation zones caused by water injection, and therefore will not readily enter the newly opened zones containing higher oil saturation.

Another consideration is the completion interval selection in bottom-water situations. There is a large volume of oil between the point of water-free oil production and 100% water production. Since waterfloods produce water, there is no reason to maintain the restricted/limited entry completion intervals that were designed to delay the onset of water coning in producing. However, the injection wells should not be completed too close to water as the potential for lost injection into the aquifer is significant.

Keep all producing wells pumped-off


Most if not all producing wells in waterfloods should be on artificial lift. Keeping the wells pumped off will maximize production. If a well is not pumped off, not only will production decrease with lower pressure zones ceasing to produce (reducing sweep efficiency), but crossflow can occur which hurts production even more.

Inject below formation fracture pressure

The main objective of injection wells is to inject the maximum amount of water without having it go out of the intended pat zone. This requires that water be injected at the highest pressure possible without fracturing/parting the formation.

Inject clean water


There are four main problems with injection water that can cause a reduction in injection capacity:

    • Dissolved solids in the injection water that can precipitate and form scale,
    • Oil and suspended solids that can plug wellbores,
    • Oxygen in the water that can cause corrosion,
    • Bacteria in the system that can cause corrosion and suspended solids.

Cleanup of injection water for oil and suspended solids can be accomplished either by filtration or by providing sufficient settling time in tanks.

Tanks, injection pumps, and injection lines are subject to scaling if make-up water and produced water have scaling tendencies and are mixed prior to injection. Splitting the injection system is the best solution for preventing scaling in the surface injection facilities, i.e. have a tank and pump system for both make-up and produced water.

In carbonate systems, the injected water will dissolve minerals as it moves through the formation until it reaches saturation, and then precipitate scale as pressure drops in the producing wellbores. The only effective solution to this problem is to squeeze inhibitor into the formation periodically, but these treatments can be expensive. One work-around solution is to use brackish water as make-up water with the spilt injection system described above.

Operate the waterflood based on injection well tests

Management of the waterflood should have management of the injection wells as its primary focus. The injectors should be periodic tested as follows:

    • Step rate tests to determine current formation parting pressure
    • Temperature, spinner, and radioactive tracer surveys to determine injection profiles
    • Pressure fall-off tests to determine skin factor, fracture half-length, and reservoir pressure.
    • Pulse tests or interference tests occasionally are run in large projects to determine the pressure communication between wells and to estimate interwell rock properties.

Injection profiles identify perforations that are not taking injected water, possibly because the perforations are plugged. These profiles also may reveal injectivity contrasts within a perforation set. A failure to correct for high concentration of injection over small intervals (thief zones) can result in poor vertical sweep, lower oil recovery, and increased water production. An injection profile should be conducted every six months during the first two years of a well’s injection life, annually thereafter, but within 30 to 60 days of any remedial work. To ensure that injection into each layer is in proportion to its displaceable hydrocarbon pore volume, profile modification may be required.

Foams, polymers, selective completions, and multiple packer completions for profile control have all been used with varying degrees of success. The success or failure is often dependent on the mechanical condition of the well. Corrosion-resistant casing (Stainless-steel, fibre-glass, etc.) has been installed across injection well completion intervals.

Conduct a surveillance program


The successful management of a waterflood project relies heavily on the surveillance program. Valuable insights into the performance of the waterflood can be gained from collection and interpretation of the following data:

    • Production volumes. Testing of each producing well is essential, preferably for a minimum of 24 hours using a three-phase separator.
    • Water injection volumes and surface injection pressures.
    • Individual well plots of injection rate versus time, injection pressure versus time, and Hall plots.
    • Pattern-injection withdrawal ratio plots, and total project injection-withdrawal ratio plots.

Full-field reservoir simulation can locate bypassed oil, identify viable (and uneconomic) drilling locations, and provide guidance on optimal changes in operations and waterflood pattern alignment.

Significant improvements in efficiencies in large, multi-well patterns can be achieved through the careful management of flood injection rates in each waterflood pattern. Balancing injection and production rates within and between patterns can substantially reduce produced water, improve long-term production rates, and enhance ultimate recovery.

References

Buckley, S.E. and Leverette, M.C.: “Mechanism of Fluid Displacements in Sands”, Trans., AIME (1942) Vol. 146, pp. 107-116.

Caudle, B. and Witte, M.: “production Potential Changes During Sweep-out in a Five-Spot System”, Trans. AIME, (1959), Vol. 216, pp. 446-448.

Dykstra, H., and Parsons, R., “The Prediction of Oil Recovery by Water Flood,” in Secondary Recovery of Oil in the United States, 2nd ed. Washington, DC: American Petroleum Institute, 1950, pp. 160–174.

Gulick, K.E. and McCain, W.D., Jr.: "Waterflooding Heterogeneous Reservoirs: An Overview of Industry Experiences and Practices", Paper SPE 40044 presented at the SPE International Petroleum Conference and Exhibition, Villahermosa, Mexico March 3-5, 1998.

Interstate Oil Compact Commission, Improved Oil Recovery, Oklahoma City, 1983.

Khelil, C. (1967): ''A Correlation of Optimum Free Gas Saturation with Rock and Fluid Properties'' (preprint), SPE 1983, Society of Petroleum Engineers of AIME, Dallas, Texas, 109-115.

Smith, J.T. and Cobb, W.M.: "Predicting Waterflood Recovery Performance", Presented at the PTTC Midwest Region Workshop, Evansville, IN, February 17-21, 1997.