Sampling and analysis of produced water

The composition of subsurface water commonly changes vertically and laterally in the same aquifer. Changes may be brought about by the intrusion of other waters and by discharge from and recharge to the aquifer. As a reservoir is produced, the compositions typically change with time; therefore, it is difficult, but important, to obtain a representative sample of a given subsurface body of water. Any one sample is a very small part of the total mass, which may vary widely in composition; therefore, it is generally necessary to obtain and analyze many samples. Also, the samples themselves may change with time as gases evolve from solution or may precipitate solids when coming to ambient conditions. The sampling sites should be selected to fit into a comprehensive network to cover an oil-productive basin, in which case the information is of value in both exploration and production. This page discusses sampling and analysis of oilfield (produced) water.

Need for periodic sampling

Water compositions from particular wells change over time, so a periodic sampling program is needed. The sampling frequency needed is not universal. The best guideline is the particular field itself: the faster the composition changes, the more often the water needs sampling. Fields undergoing waterflood or tertiary recovery naturally show water-chemistry changes as the injected water mixes with the formation water. Reservoirs under primary production can show dilution from water moving from adjacent compacting clay beds into the petroleum reservoir as the pressure declines because of the removal of oil, gas, and brine. This can be a detector for possible subsidence of the reservoir and perhaps the surface. The composition of oilfield water can vary with the position within the geologic structure from which it is obtained. In some cases, the salinity will increase upstructure to a maximum at the point of oil/water contact.

Water sampling methods

Various techniques and devices are now available for the evaluation of newly drilled wells, but there remain some important factors that must be considered when obtaining these samples and evaluating the results. The primary problem stems from the contamination of the reservoir with the drilling, completion, and stimulation fluids used during the well-construction operation. The contamination is usually easily detected from the analytical results. If one realizes that high concentrations of particular species that were used in those wellbore fluids appear in the water-analysis results, then the sample was likely contaminated and the usefulness of the sample reduced accordingly. Naturally, one must always include these species in the analysis. Unfortunately, if a sample is contaminated there is seldom an opportunity to re-enter the hole and obtain a new sample, except at great cost.

One approach for a standard drillstem test (DST) is to sample the water after each stand of pipe is removed. Normally, the total-dissolved-solids (TDS) content will increase downward, becoming constant when pure formation water is obtained. A test that flows water improves the chances of an uncontaminated sample; otherwise, the best chance is to sample just above the DST tool, because this was the last water to enter the tool. Newer downhole samplers sometimes allow multiple samples to be obtained; therefore, if one first pulls a large volume of water into the tool and then a second (or third) sample from the same interval, there is a chance of getting uncontaminated water in that sample.

For example, analyses of water obtained from a DST of Smackover limestone water in Rains County, Texas, demonstrates the errors caused by improper sampling of DST water. Analyses of top, middle, and bottom samples taken from a 50-ft zone of fluid recovery show an increase in salinity with depth in the drillpipe, indicating that the first water was contaminated by mud filtrate. Thus, the bottom sample was the most representative of Smackover water.

Sample containing dissolved gas

Knowledge of certain dissolved hydrocarbon gases is used in exploration. Mapping anomalies of hydrocarbons in both surface water and subsurface aquifer water samples is an extraordinarily powerful geochemical tool.

Sampling at the flowline

Another method of obtaining a sample for analysis of dissolved gases is to place a sampling device in a flowline, as Fig. 1 illustrates. The device is connected to the flowline, and water is allowed to flow into and through the container, which is held above the flowline, until 10 or more volumes of water have flowed through the container. The lower valve on the sample container is closed, and the container removed. If any bubbles are present in the sample, the sample is discarded, and a new one is obtained.

• 

  Fig. 1 – Flowline sampler.

Sampling at the wellhead

It is common practice to obtain a sample of formation water from a sampling valve at the wellhead. A plastic or rubber tube can be used to transfer the sample from the sample valve into the container (usually plastic). The source and sample container should be flushed to remove any foreign material before a sample is taken. After flushing the system, the end of the tube is inserted into the bottom of the container, and several volumes of fluid are displaced before the tube is removed slowly from the container and the container is sealed. Fig. 2 illustrates a method of obtaining a sample at the wellhead. An extension of this method is to place the sample container in a larger container, insert the tube to the bottom of the sample container, allow the brine to overflow both containers, and withdraw the tube and cap the sample under the fluid.

• 

  Fig. 2 – Example of the method used to obtain a sample at the wellhead.

At pumping wellheads, the brine will surge out in heads and be mixed with oil. In such situations, a larger container equipped with a valve at the bottom can be used as a surge tank, an oil water separator, or both. To use this device, place the sample tube in the bottom of the large container, open the wellhead valve, rinse the large container with the well fluid, allow the large container to fill, and withdraw a sample through the valve at the bottom of the large container. This method obtains samples that are relatively oil free.

Field-filtered sample

For some studies, it is necessary to obtain a field-filtered sample. Fig. 3 shows a filtering system that has proved to be successful for various applications. This filtering system is simple and economical. It consists of:

• 50-mL disposable syringe
• Two check valves
• Inline-disk filter holder

The filter holder takes size 47-mm-diameter, 0.45-μm pore-sized filters, with the option of a prefilter and depth prefilter. After the oilfield brine is separated from the oil, the brine is drawn from the separator into the syringe. With the syringe, it is forced through the filter into the collection bottle. The check valves allow the syringe to be used as a pump for filling the collection bottle. If the filter becomes clogged, it can be replaced in a few minutes. Approximately 2 minutes are required to collect 250 mL of sample. Usually three samples are taken, with one being acidified to pH 3 or less with concentrated HCl or HNO₃, and another stabilized with biocide. The system can be cleaned easily or flushed with brine to prevent contamination.

• 

  Fig. 3 – Example of field-filtering equipment.

Sample for stable-isotope analysis

Stable isotopes have been used in several research studies to determine the origin of oilfield brines. The most common isotopes studied are deuterium/hydrogen, ²H/¹H; oxygen, ¹⁸O/¹⁶O; and stable carbon isotopes, ¹³C/¹²C. Also, isotopically labeled compounds are sometimes used to trace injection waters or stimulation fluids. The isotopic analyses of the water, gas, and oil phases can be useful to geochemists in determining the sequence of fluid migrations in reservoirs during their genesis.

Sample for determining unstable properties or species

A mobile analyzer was designed to measure:

1. pH
2. Eh (redox potential)
3. O₂
4. Resistivity
5. S⁼
6. HCO^3-
7. CO³⁼
8. CO₂

Portable field kits for chemical analysis are available for these analyses, which provide reasonably accurate on-site results. When oilfield brine samples are collected in the field and transported to the laboratory for analysis, many of the unstable constituents change in concentration. The amount of change depends on the following:

• Sampling method
• Sample storage
• Ambient conditions
• Amounts of the constituents

Therefore, an analysis of the brine at the wellhead is necessary to obtain reliable data.

Sample containers

The types of containers that are used include polyethylene, other plastics, hard rubber, metal cans, and borosilicate glass. Glass will adsorb various ions, such as iron and manganese, and may contribute boron or silica to the aqueous sample. Plastic and hard-rubber containers are not suitable if the sample is to be analyzed to determine its organic content, unless tested and shown otherwise. A metal container is used by some laboratories if the sample is to be analyzed for dissolved hydrocarbons such as benzene. Produced water often will corrode metal containers, unless they are lined. The corrosion will alter the chemical composition of the water by adding metal ions such as iron and manganese and by lowering the pH. The type of container selected depends on the planned use of the analytical data.

Water sampling and analysis specification

This section defines a "standard water analysis" that specifies the complete suite of species and provides suggestions on sampling and analytical techniques. The analysis specifications have been used by the author for two decades and have been used by others, with adaptations for particular needs. The results of this extensive list of analyses are the required input for the various computer multicomponent simultaneous-equilibrium chemistry models that can predict correctly the changes in the:

• Water chemistry
• Scale precipitation
• Rock/water interactions
• Water sources and mixing
• Corrosion

While there is always temptation and pressure to limit the range of analyses performed on a sample, analytical technology has reduced dramatically the value of limiting the range of an analysis. Most instruments automatically measure many different species in a single run; therefore, it may cost the same to get one species as it does to get many, depending on the analytical technique. For instance, the commonly used ICP instruments typically measure 10 to 60 elements at a time (depending on the particular instrument) in a few minutes; therefore, adding or subtracting a species on the ICP changes the actual cost very little.

The analytical results should be stored in an electronic database for further analysis and comparison of samples from the same points over time. Certain evaluations can be performed on individual samples, such as scale-deposition predictions or monitoring for corrosion with dissolved manganese and iron. Other evaluations require successive samples, with consistent sample-collection and -analysis procedures, for the data to be meaningful. These results can be mapped to follow the reservoir-area changes over time. They also can provide insight as to potential improvements in reservoir management of a waterflood, because the various ions provide natural tracer data for the injected water. The data are also useful for deciphering casing leaks, conductive (but unmapped) faults, and, in favorable cases, zonal splits of water production. If the data were not collected on each sample consistently or if the analysis suite was minimized to save money in the laboratory, these valuable evaluations are not possible.

Fig. 4 illustrates the information that should be obtained for each sample of oilfield water. As much information as possible should be recorded when taking the sample. The following measurements must be taken in the field at the sampling point coincident with taking the sample because these values change quickly once the sample is removed: temperature, pH, Eh (optional), dissolved O₂ (Chemet or meter), H₂S, CO₂, pressure, and bicarbonate (HCO₃₋) alkalinity titration (recommended).

• 

  Fig. 4 – Description and label for each water sample.

This stabilizes the samples so that precipitation of the metals does not ruin the results for the ICP analysis. In particular, the iron will precipitate as the sample sits and that precipitate removes many of the other species from solution. Also, the naturally present bacteria in the water will quickly eat the organic acids in the samples unless biocide is added to that sample. If no biocide, such as glutaraldehyde, has been added, no organic acids are detected; therefore, adding glutaraldehyde to a bottle used for the organic-acid analysis is highly recommended.

In addition to the water analysis, the gas in contact with the water is important. Taking a gas sample at the same time is recommended, although one taken at a different time can suffice for less critical applications. The main constituents of interest in the gas are the CO₂ and H₂S. The H₂S analysis usually is done at the sampling point. CO₂ and H₂S are very important in predicting what will happen with the water.

In general, bacterial surveys are recommended, primarily with the serial dilution technique with one medium for sulfate-reducing bacteria (SRB) and another medium for general anaerobic bacteria. Many times, the exact medium composition (and sometimes the incubation temperature) has to be adjusted for a particular field to get that population to grow well. The bacteria have a lot of negative side effects, such as:

• Generating H₂S
• Causing a safety and corrosion problem
• Making the schmoo (organic/inorganic deposits) formation in the produced-water problem worse.

Microbial-induced corrosion can be very severe, especially in :

• Deadlegs (stagnant areas in which water and solids are trapped)
• Low-flow areas
• Underdeposits

Example problem

Problem. Conduct laboratory analyses to generate the data needed to enter into the computer programs that will predict the stability of water, with respect to precipitation of scales as a function of temperature and pressure.

Solution. Obtain three bottles for each sample: one for cations/metals (Table 1), one for anions/etc. (Table 2), and one for organic acids (Table 3). The bottles should be polyethylene, Teflon, or polycarbonate.

Not all polyethylene bottles are satisfactory because some contain relatively high amounts of metal contributed by catalysts in their manufacture. Fill a bottle with acid solution, let stand for a week, and then analyze the acid for metals to test the bottle’s suitability. Glass should be avoided if the sample can freeze, because it is likely to break. Glass often is reactive with highly saline brines and strongly adsorbs some oils and other organics, including treating chemicals.

• 

  Table 1 - Water sampling of bottle A (acidified) for cations/metals.

• 

  Table 2 - Water sampling of bottle B (no additives) for anions, etc.

• 

  Table 3 - Water sampling of bottle C (biocide) for organic acids.

Filter a measured portion of the sample taken for the anion analysis through a 0.45-μm filter to measure the total suspended solids. If a large amount of solids is collected, the filter paper should be saved for later X-ray powder diffraction, scanning electron microscopy/energy dispersive X-ray spectroscopy analysis (to identity the solids), and a laser particle-size analysis.

Tables 1 and 3 specify that the bottles for the cation and organic-acids analyses have additives:

• 10 mL of ultrapure nitric acid (HNO₃) for the ICP analysis
• 1 mL of glutaraldehyde for the organic acids

This stabilizes the samples so that precipitation of the metals does not ruin the results for the ICP analysis. In particular, the iron will precipitate as the sample sits and that precipitate removes many of the other species from solution. Also, the naturally present bacteria in the water will quickly eat the organic acids in the samples unless biocide is added to that sample. If no biocide, such as glutaraldehyde, has been added, no organic acids are detected; therefore, adding glutaraldehyde to a bottle used for the organic-acid analysis is highly recommended.

Noteworthy papers in OnePetro

Use this section to list papers in OnePetro that a reader who wants to learn more should definitely read

External links

Use this section to provide links to relevant material on websites other than PetroWiki and OnePetro

See also

Produced oilfield water

Produced water properties

Dissolved constituents in produced water

Suspended solids in produced water

Mixing of produced water

PEH:Properties_of_Produced_Water

Page champions

E. Dwyann Dalrymple