SECTION 4
DEALING WITH WATER PRODUCTION DURING WATERFLOODS

Water Injection and Production Trends.  One basic concept to successful waterflooding is getting the water to where the oil is. In many instances this requires converting the best producing wells to injection wells. Some independent producers select poor or marginal producers as injection wells for economic reasons and later wonder why the waterflood did not perform as expected.

In planning a waterflood: 1) Determine water requirements as accurately as the data permit; 2) Survey all possible water sources, with special attention given to satisfying quantitative requirements; and 3) Develop the selected source in the most economic manner permitted. The largest daily demand for water occurs during the fill-up period when there is no return water available. During this fill-up period, it is usually advantageous to maintain a high rate of injection, so as to accomplish an early fill-up (a rate between 1 and 2 B/D/acre-foot is desirable). After fill-up has been achieved, a rule of thumb is injection rate should be about 1 B/D and not less than ½ B/D/acre-foot.¹¹ Flood pattern, well spacing, and injection pressures should be designed to meet these requirements.

The pore volume (PV) method has been found to give a good approximation of the ultimate water requirements for a waterflood. The volume of water required should range from 150 to 170% of total pore space, and the measurement of such space should include the PV of any adjacent overlying gas interval or basal water zone.¹¹ The ultimate water requirements, together with the average water injection rate, will serve as a basis for estimating the total life of the waterflood.

If gas or water intervals are not present, produced water will comprise 40 to 50% of the ultimate water requirements. If gas and water intervals are present, less return water will be available – thus, the ultimate make-up water requirement will increase to as much as 60 to 70% of the total quantity of water injected.

The volume of produced water will increase during the life of a waterflood, both from an individual well and overall project basis. As the flood front reaches a producing well, fluid volumes will increase, making it necessary in most instances to increase the capacity of the artificial lift equipment. It is important to capture (produce) as much of the oil as possible as the flood front advances past the producing well. Fluid level monitoring will help insure the producing wells are being pumped off.

As the flood front advances past the producing well, an increasingly higher percentage of water will be produced. In many instances it is advantageous to shut in the producing well or convert it to an injection well when it reaches a high water cut, in order not to rob water from the advancing front. If it is shut in, it can be reactivated later and produced to an economically limiting water-oil ratio.

Material balances should be run between injection and production wells. Monitoring the amount of water injection and the amount of water production can lead to important information. This is a comparison of water in versus water out. Do this comparison periodically for the project and for areas or patterns within the project. This is sometimes referred to as pattern water balancing. The comparison can be accomplished using injection/withdrawal ratios.

If the volume of injection water into an area or pattern is similar to the amount produced in a nearby well, this could be an indication of a channeling or communication problem. Simple tests, such as stopping or decreasing water injection and monitoring changes in the amount of water produced, can help detect a problem. If a correlation exists, then a tracer test should be run to verify the problem.

Reinjecting Produced Water Versus Make-up Water. The volume of return water becomes an increasing percentage of the required injection rate as a flood progresses; therefore, it is an economic necessity that produced water be reinjected unless the cost of treating the produced water is higher than that of the make-up (or source) water. Incompatibilities between different waters and the reservoir rock must be considered. Pore-size distribution, and composition and distribution of clays are the most important rock considerations. Complete water analysis of both the injection source water and connate water enables scaling, clay-swelling, and brine incompatibilities to be evaluated. Bacteria, suspended solids, oil, and dissolved oxygen and hydrogen sulfide levels should be established. Special attention should be given to the detection of any combinations of ions that may precipitate on being mixed. Unacceptable levels of these parameters must be addressed in the facility design and chemical treating programs. In many major waterfloods, waters are isolated in the surface system and are injected separately into the reservoir. When mixing incompatible brines cannot be avoided, they should be mixed on the surface rather than downhole so that resulting scales or deposits can be more easily removed. Chemical treating requirements, backwashing and acid treatments on injection wells will increase when incompatible brines are mixed.

Water Quality Needed to Maintain Injectivity. The proper balance between water quality and cost must be determined. While rigorous water quality guidelines like 98% removal of particles above 2 microns, oil < 5 ppm, and oxygen < 50 ppb¹² will nearly always provide water of sufficient quality, costs to achieve that quality can become excessive. Cost elements include initial installation costs of water treating facilities, chemical costs, frequency and cost of well cleanup workovers, and other maintenance and operating costs. The economic analysis must consider delayed production due to poor injectivity and potential lost production due to reduced sweep efficiency. Water handling facilities must be designed to handle treating upsets, since even short periods of high oil carryover can cause significant formation damage in injection wells.

In the past, fresh water was commonly used in waterfloods. Because of increasing scarcity, fresh water will not generally be a viable source. Therefore, most injection projects use saline or brine waters.  For a waterflood operation to be successful, the water used for injection must be of a quality that will not damage the reservoir rock and injection rates must be maintained below the parting pressure of the reservoir. Poor water quality will result in lost oil production.

Five components in water detrimental to a waterflood are: 1) microorganisms, 2) dispersed oil, 3) suspended solids, 4) dissolved gases, and 5) dissolved solids.¹³

Three classes of microorganisms found in water used in the oil field are algae, fungi, and bacteria, with bacteria representing the most serious problem in waterfloods. They range in size from 0.2 to 10 microns. Bacteria are controlled using biocide chemicals and removed by filtration.  

Dispersed oil in injection water is detrimental for three reasons. First, bacteria utilize certain components in the crude oil as food. Second, oil is strongly adsorbed on iron sulfides and other scale deposits, which makes it difficult to remove these deposits with acid treatments. Third, oil reduces the relative permeability to water in the injection well. As relative permeability to water decreases, it requires more pressure to inject the same amount of water. Dispersed oil in injection water can be reduced by proper use of demulsification chemicals and by better design of the water system.

Suspended solids are either organic, from algae and bacteria, or inorganic, from minute particles of clay and sand or precipitates of calcium carbonate, iron sulfides, and other scales. Many can be removed by settling tanks and filters and must be removed in order to prevent damage to the injection well. Complete removal of all suspended solids is difficult and expensive, especially for very small particles below a micron in size. As a rule of thumb, all particulate matter larger than one-third the average pore throat diameter for the reservoir for which the water is intended should be removed. The average pore throat diameter, in microns, can be estimated by taking the square root of the formation permeability in millidarcies. For example, if the formation permeability is 100 millidarcies, then the average pore throat diameter is estimated at 10 microns. Thus, the injection water should have no particulate matter greater than 3.3 microns.

Dissolved gases frequently found in injection waters are oxygen, carbon dioxide and hydrogen sulfide. All three enhance corrosion problems. Oxygen can be removed by an oxygen scavenger, such as cobalt-catalyzed sodium bisulfite. Proper gas blanketing of water tanks also minimizes oxygen entry. Hydrogen sulfide can be oxidized to sulfur with oxygen or sulfur dioxide, or to sulfate with hypochlorite. Removal of carbon dioxide from the water can be achieved by stripping with an inert gas, such as nitrogen, but the cost generally exceeds the benefit.

Dissolved solids are found in all waters. Common materials found in oil field waters include sodium, calcium, magnesium, barium, strontium, ferrous and ferric, and aluminum cations, along with carbonate, bicarbonate, sulfate, sulfide, chloride, bromide, iodide, and silicate anions. Have your water analyzed on a regular basis and implement a chemical treatment program to help minimize problems.

Using Chemical Tracers to Identify Channeling or Premature Water Breakthrough. A chemical tracer is a tool to determine water flow from an injection well to surrounding production wells. Tracing injected water with a chemical and observing when and where that chemical is produced can provide information on directional flow trends, identification of rapid interwell communication, volumetric sweep, and delineation of flow barriers.

Water-soluble chemical tracers added to the injection water stream are substances not normally present in formation fluids. Ideally, these tracers do not interact with either the reservoir rock or other fluids in the reservoir. Injection water movement is monitored by analyzing produced water in area wells for the presence and concentration of the tracer. It is important to monitor wells beyond the immediate offset producers. Tracers can be injected as a high concentration slug or continuously over a longer time period. When the slug method is used, offset producers must be sampled every few minutes or hours to detect a tracer spike as the slug flows by. With the continuous method, sampling can be less frequent. Inferences concerning channeling and areal sweep efficiency are then drawn from analysis of tracer transit times between injection well and producers and from concentration levels observed in these wells.

In most instances, before any tracers are injected, the reservoir should be “pressured up.” This means the reservoir must be on waterflood long enough to fill void spaces to minimize loss of tracer material.

Common tracers are fluorescein sodium dyes, ammonium nitrate or fertilizer, ammonium thiocyanate, and lower molecular-weight alcohols such as methanol and isopropanol. Fluorescein sodium dyes generally are used only when severe channeling with short residence times is suspected because adsorption by the reservoir rock can be excessive over longer periods of time. Fluorescein sodium dyes can be visually detected at very low concentration levels, making field detection very simple and inexpensive. Like the fluorescein sodium dyes, ammonium nitrate is inexpensive and can be detected in the field, but its use requires specific reagents and colorimetric equipment. Ammonium thiocyanate can be used in applications similar to ammonium nitrate, but it costs more and is not always available. Limitations on the use of lower molecular-weight alcohols include higher cost and the need for laboratory analysis with gas chromatograph equipment.

Using Gelled Polymers to Modify Permeability to Increase Sweep Efficiency. Many waterfloods are plagued with low volumetric sweep efficiency. In many instances, poor performance is thought to be a result of water moving rapidly through high permeability channels or through natural or induced fractures. Induced fractures are often the result of over-pressuring the formation at some point. In other instances, water breakthrough may be related to permeability contrasts between different layers, which may or may not be in vertical communication in the reservoir.

Permeability modification treatments can help improve volumetric sweep efficiency. In waterfloods, injection-side treatments are most common. These treatments are conducted with either crosslinking or in-situ polymerization processes.

Crosslinked polymer treatments involve the addition of low concentrations of metal ions to the polymer solution causing the polymer molecules to bond to one another, greatly increasing the resultant gel’s ability to develop resistance to the flow of fluids in the reservoir rock. Depending on the concentration of polymer, crosslinking agent and rate of combining the two, a wide range of permeability adjustment is possible. In the in-situ polymerization process, monomers are polymerized in the reservoir. When treatments are properly placed in the targeted area, resulting fluid flow changes in most cases improve oil recovery and reduce operating costs due to reduced water cycling. Long-term performance of these treatments relies on the in-situ solutions having sufficient strength to stay in place during drawdown.

Permeability modification treatments must address correct identification of geological and reservoir characteristics, correct design and effective placement in the reservoir, and effectiveness lasting throughout the project period. Since different conformance improvement technologies are not applicable to all reservoir problems, the critical task is to successfully identify the channeling problem and then to match an appropriate technology to that problem.

Past historical success rates with crosslinked polymer treatments have been less than 50%¹⁴, with more than two-thirds of failures resulting from improper application¹⁵. Most failures with in-situ gel polymer are caused by one or more of the following: 1) improper placement of gel polymer within the wellbore, 2) improper selection of candidate well, 3) lack of knowledge of wellbore integrity, 4) lack of adequate preparation of wellbore prior to the job, 5) limited time allotted to implement the treatment, and 6) not understanding the injection well profile prior to and after the treatment. With proper engineering, planning, and application, success ratios of more than 80% have been achieved.¹⁴

Facts operators should know about gelled polymers: 1) Gels are created when dry polymer is mixed in water and crosslinked with a metal ion (usually chromium triacetate or aluminum citrate); 2) Gelation time is controllable ranging from a few hours to weeks; slower gelation time allows for more volume and deeper placement; 3) Gels having viscosity and elasticity ranging from slightly greater than fresh water to rubber can be created in virtually any water, at temperatures up to 400^o F, in high H₂S environments; 4) Special equipment is normally required to properly blend and pump polymer gels; 5) Gels can be created that completely block the flow of fluid through all reservoir rock or they can preferentially reduce permeability and fluid flow through only the most permeable and conductive pathways; 6) Gels can be created with polymer concentration ranging from a few hundred to more than 50,000 ppm; low polymer concentration means less gel strength and higher concentration means more gel strength; 7) Weaker gels (colloidal dispersion gels) are used in reservoirs dominated by matrix flow conditions and stronger gels (bulk gels) are used in reservoirs dominated by fracture or vug flow conditions; 8) Gels are equally applicable to sandstone and carbonate reservoirs; and 9) Gels are relatively inexpensive because they contain 98% or more water.¹⁶

Candidate well selection. Selection criteria for injection well candidates are: 1) significant remaining mobile oil-in-place that can be recovered if sweep efficiency is improved, 2) low secondary oil recovery due to poor sweep efficiency (i.e., high degree of reservoir heterogeneity), 3) premature water breakthrough at producing wells, 4) evidence of direct injector to producer channeling through fractures, vugs or high matrix permeability rock, and 5) high injection rate associated with low wellhead pressure.

Two important factors in determining waterflood efficiency are reservoir heterogeneity and mobility ratio. Mobility of a fluid is defined as the effective permeability of a particular fluid divided by its viscosity. In a waterflood, the mobility ratio is the mobility of the displacing phase (water) divided by the mobility of the displaced phase (oil). For mobility ratios greater than 1.0, polymer-augmented flooding should be investigated.

Reservoir heterogeneity can also be improved using gelled polymer treatments. Reservoir heterogeneities are nonuniformities in reservoir properties that cause inefficient sweep and premature water breakthrough in a waterflood. These heterogeneities depend upon the depositional environment of the reservoir; subsequent events such as compaction, solution, cementation, and fracturing; and the nature of the particles constituting the reservoir. Porosity and permeability are reservoir properties that often exhibit significant heterogeneity in both the vertical and areal sense, but they are generally more pronounced in the vertical direction.

The degree of heterogeneity can be determined from cores. Core reports can be used to predict performance of new floods and to explain reasons for poor performance in mature floods. The Dykstra-Parsons coefficient of permeability variation is a common descriptor of reservoir heterogeneity. It measures reservoir uniformity by the dispersion or scatter of permeability values. A homogeneous reservoir has a permeability variation that approaches zero, while an extremely heterogeneous reservoir would have a permeability variation approaching one. If a core analysis is not available, a nearby core from the same reservoir can be used as an approximation.

When to use polymer gels in injectors. Inject colloidal dispersion gel at the inception of a waterflood to avoid sweep efficiency problems if performance of an analogous flood suggests that premature water breakthrough will be a problem, or representative core data indicate that the reservoir has a high Dykstra-Parsons coefficient (greater than 0.6) and will not flood uniformly. Inject bulk gel after waterflood inception if water channeling through fractures or high permeability streaks creates a sweep problem. Expected results are increased resistance at the injector, more oil produced faster and at lower water-oil ratios, and less water to handle.

Treatment design. The initial step in treatment design is selecting a process appropriate for the reservoir/producing problem and the treating/reservoir fluids. Choices to be made include near-wellbore versus deep gel treatments, type of polymer, crosslinking agent, and crosslinking process. On-site and laboratory testing by service companies with actual treating/reservoir fluids assists in chemical selection and treatment design.

A critical step is calculation of treatment volume and prediction of variation in polymer composition. Diverse tools, such as production/injection histories, well logs, surveys, workover history, and personal knowledge of the formation and geographical area are critical for prediction of treatment volume. It is impossible to calculate treatment volume exactly, but estimation within reasonable limits is possible. That is why it is essential that injection rate and pressure be continuously monitored during treatment and appropriate changes made to optimize treatment.¹⁷  Polymer solution should be injected until parting pressure is approached while injecting, the injected slug is produced at a peripheral producer, or the maximum design size is achieved. In most cases, parting pressure is the limiting factor for treatment size.

Interwell chemical tracer data provide valuable information for designing the size, gel time, and gel strength.¹⁸ Tracer breakthrough time and injection rate can provide an estimate of the size of the channel and, thus, the treatment size. In utilizing this technique, it is important to know the concentration of the tracer showing up at the producing well. If it is much lower than the injected concentration, it might be wise to increase the treatment size proportionally. Usually more than the fracture volume of polymer must be injected, since it will leak into the reservoir matrix. Tracer breakthrough time can also provide an estimate of the required gelation time and gel strength. For in-depth placement of the gel solution, the gelation time should be long enough to place the entire treatment.

Tables 3 and 4 provide examples of the volume of gelling solution needed to fill fractures of various dimensions.

Table 3.  Estimated Treatment Volume of a Single Horizontal Fracture Radial Width 0.05 in

Table 4.  Estimated Treatment Volume of a Single Vertical Fracture Width 0.05 in; Height 50 ft

Performing the treatment. The candidate well should be cleaned prior to pumping the polymer solution. The goal of wellbore cleaning, whether mechanical or chemical, is to provide a clean formation face free of sludges, solids, or other materials that might interfere with the polymer solution or affect injectivity. Chemical performance and compatibility should be checked in the actual fluids on-site, since trucks and frac tanks can be sources of contaminants. Mixing/injection procedures must ensure that uniform polymer mixes are prepared.

Design treatment volumes and chemical concentrations should be used only as guidelines. Since each well will have a unique response, the well’s ability to accept fluid (injectivity) should be continuously evaluated during treatment, and treatment compositions adjusted accordingly or the treatment terminated before the design volume is injected if the injection rate decreases too much. Rate restriction to avoid formation parting is essential. Hall plot slope analysis is very useful for real-time monitoring of treatments. The summation of surface or, preferably, bottomhole pressure multiplied by time versus cumulative treatment volumes are plotted at 15-minute intervals during treatment, and slopes determined. Changes in slope can indicate if treatment should be terminated before the design volume is injected.

Other recommendations for placing a gel treatment are: 1) increase polymer concentration in stages, 2) inject treatment at a rate similar to normal injection rate, 3) stay below reservoir parting pressure, 4) keep offset producers active during treatment, and 5) over-displace the treatment with water (use more water in short perforated intervals).