Generating Electricity From Produced Water
There are 823,000 oil and gas wells in the United States that coproduce hot water with their hydrocarbon output. This equates to approximately 25 billion bbl of water annually that could be used as fuel to produce up to 3 GW of clean electrical power. Not only would electricity generated from produced water add much needed electrical power, the life of many of these wells also would be extended and additional oil and gas produced.
A recent project funded by the National Energy Technology Laboratory of the United States Department of Energy conducted field demonstrations to determine the potential of generating electricity from hot produced water. Participants included Gulf Coast Green Energy (GCGE), ElectraTherm, Denbury Resources, the Southern Methodist University (SMU) Geothermal Laboratory, the Texas A&M University petroleum engineering department, and Dixie Electric Cooperative.
The primary goal of the project was to prove the feasibility of interfacing the ElectraTherm Green Machine, a waste heat-to-power generator, with a producing oil or gas well. The project’s subsidiary goals were as follows:
- Demonstrate the ability to produce electricity from waste heat in the produced water.
- Show that producing electricity from produced water does not interfere with normal well operations.
- Demonstrate how small oil and gas producers might increase their profitability by adding an income stream from power generation.
- Determine the economic viability of generating electricity from waste heat in the produced water.
- Determine whether the kWh output of electricity from the produced water has practical applications.
- Identify the environmental impact of generating fuel-free, emission-free electricity from waste heat in the produced water.
It was important to hold a field trial to determine the extent of known factors that could not be identified or quantified in laboratory and bench scale runs. In addition, a field trial was necessary to identify potential corrective measures for new equipment designs and future produced water projects. The site chosen was a producing oil well, Denbury’s Summerland No. 2 well, near Laurel, Mississippi. In production for 5 years, the well has a high water cut and high produced water temperature. The well produces 100 BOPD and 4,000 BWPD from a depth of 9,500 ft with an electric submersible pump. The temperature of the produced water exiting the “knockout” tank at 120 gal/min is 204°F. The site has an ambient temperature range of 60°F to 105°F.
Choosing the Right Technology
Organic Rankine Cycle (ORC) generators create pressure by boiling various chemical working fluids (refrigerants) into a high-pressure gas. The gas expands in a one-way system and turns an expander or high-speed turbine, which drives a generator that generates electricity. Historically, ORC generators incorporating turbo expanders or turbines have not been commercially viable in sizes less than 1 MW. However, the Green Machine uses a patented, robust, low-cost twin screw expander (see Fig. 1) that requires much less water volume than the larger ORC generators. The Green Machine can generate between 30 kWh and 65 kWh with hot water flows of 200 gal/min or less, and because most oil and gas wells produce less than 200 gal/min of hot water, it was selected for this demonstration. The generator was also chosen for its relatively small size and portability. It is skid mounted, can be moved with a small forklift, and has a minimal equipment footprint of 300 ft2.
While the technology is relatively new, a prototype unit suitable for oil and gas applications was tested and demonstrated in a boiler room application beginning in May 2008 at SMU.
When the generator is in use, produced water from the well enters a heat exchanger where the hot water excites (pressurizes) the working fluid—an EPA-approved nonhazardous, nontoxic, and nonflammable fluid—which drives the twin screw expander (the power block) to create electricity. The twin screw expander is unique in its configuration, lubrication, and specifications, but uses reliable, proven compressor technology that has existed for more than 20 years. The twin screw expander has a rotational speed of 4,300 to 4,800 rev/min, one-tenth of that of most turbo expanders. The robust screw allows wet vapor to travel through the expander, thereby enabling access to lower temperature resources. A patented process and lubrication scheme simplifies and/or eliminates lubrication reservoirs, oil coolers, pumps, lines, and filters, creating a simple and efficient system with fewer parasitic loads.
After the working fluid expands across the twin screw expander (spinning a generator) the low-pressure vapor must be condensed to a liquid to begin the cycle again. The condenser for the demonstration generator was air cooled to eliminate the extensive use of fresh water and the maintenance expenses associated with operating a cooling tower. The generator’s control system is fully automated, thus allowing remote control, remote monitoring, and offsite diagnostics and trending.
The generator (Fig. 2) and condenser were tested at the factory, mounted on a drop-deck flatbed trailer, and trucked to the site, where a test run was completed soon after arrival. A hot water bypass valve was installed by GCGE and Denbury field employees, which allowed the produced water to bypass the generator during downtime. Denbury laid and connected the pipe from the hot water bypass to the trailer and the final connections to the generator were made up with high-pressure hoses.
Dixie agreed to “net meter” the electricity generated by the unit and credit the electrical production at retail rates, which meant that the generated electricity was allocated directly to the field.
The 6-month demonstration successfully concluded in November 2011, with 1,136 total runtime hours, and provided excellent insight for future installations. The high summer temperatures reduced the temperature differential between the hot water temperature and the condensing temperature so much that the equipment was programmed to shut down when the ambient temperature was above 92°F. Future shutdowns could be avoided by using larger condensing fan units. The larger condensers could have added up to 40% more KWh output by increasing the heat transfer surface area for the refrigerant, thus allowing the temperature differential to increase. Because the hot water bypass valve became clogged and required a replacement valve to be installed, it was determined that the bypass valve used for produced water applications must have a different design.
Geothermal Brine Issues
Water corrosion and mineral buildup in the ORC generator’s heat exchangers were considered a major challenge going into the demonstration. The investigators understood that brazed-plate heat exchangers are not optimally suited for brine as they tend to clog and experience stress corrosion cracks. Thus, the investigators concluded that the chosen heat exchanger design would be insufficient for long-term operation. However, a 6-month-long, 1,000-hour test run using the installed heat exchangers presented no problems. The addition of a gasketed plate-and-frame heat exchanger would allow various metallurgy options and cleaning ability, as well as extend heat exchanger life. The use of a small metering pump to add a scale inhibitor to the produced water before it enters the generator is another potential solution.
A review of the demonstration and subsequent cost analysis confirmed the economic benefits of the application. A post-project analysis concluded that the Green Machine’s power generation offset about 20% of the energy required to run the downhole pump on the oil well. Thus, it provided an attractive payback at oil and gas sites where power costs more than USD 0.08/kWh and where producers see generation of waste heat electricity as a public relations or corporate social responsibility value.
For wells with increased produced water flow and/or temperature, the internal rate of return (IRR) and net revenue will be substantially greater. For example, a single well that produces 65 kWh of electricity by using the Green Machine would have an IRR of 25% with a 20-year power cost of USD 0.028/kWh and net revenue of USD 1.16 million over the life of the equipment. This will provide the incentive for oil and gas producers to continue operating a well long after it would usually be shut in because of the produced water level. It may also offer an incentive for operators to consider bringing wells onto production that previously would have been uneconomic because of their projected water volumes.
The lessons learned from the demonstration at Denbury’s Laurel site provided insight that could help future applications of the ORC generator technology to reduce installation time, increase efficiency, generate additional power, and minimize maintenance. This kind of cogeneration can be effective in reducing the energy costs particularly for pumping geographically remote oil wells, a need increasingly seen in the US.
However, some hurdles remain in developing opportunities for the use of coproduced fluid to generate electricity. Economics will play a critical role in the growth of this sector. Depending on criteria, there is an attractive return on investment in locations where the cost of power is USD 0.10/kWh or higher. In locations where the cost of power is less than that, more incentives or corporate objectives will be necessary to justify a project.
The waste heat-to-power generation technology demonstrated contributes directly to the reduction of harmful atmospheric emissions. The total electricity production was 19,180 kWh, which is equivalent to offsetting 172 tons of produced CO2. Applying the lessons learned from the air-cooled condenser, a well with a produced water flow of more than 150 gal/min generating a net output of 38 kWh of electricity would offset 360 tons of CO2 production, according to an online CO2 emissions calculator.
The process also conserves water, a focal point in major US producing regions that have recently experienced extreme drought. Operators become better resource stewards by generating electricity in the field to offset some of their electrical power use and reduce the likelihood of power shortages in an expanding electricity market.
Mapping the Way to Optimized Production in Shale Formations
A two-step analysis can provide the key information needed to design optimal shale completions. The first step is to evaluate reservoir quality, which describes the hydrocarbon potential of a shale. The second step is to evaluate completion quality, which describes stimulation potential. Core analysis provides the basis to help calibrate the results of these two steps. The intersection of good reservoir quality and good completion quality leads to the best chance for success in shale completion. However, a failure to address both reservoir quality and completion quality will jeopardize the achievement of the ultimate goal: optimized production.
A shale reservoir by definition is a hydrocarbon source, reservoir, trap, and seal in a single package. Though similar in outward appearance, no two shales are alike. They are typically complex, heterogeneous rocks with extremely low permeability. Stress anisotropy is commonplace. This calls for the judicious integration of geology, petrophysics, geomechanics, and reservoir engineering to solve the puzzle that will enable the reservoir to yield its prize.
To determine reservoir quality, defined as the hydrocarbon potential of a shale, it is necessary to quantify the amount of hydrocarbon in place and its deliverability to the fracture face. To do this, we must know the organic matter content and type, its thermal maturity, the effective porosity, fluid saturations, matrix permeability, and reservoir pressure.
The hydrocarbon in shale has evolved from thermogenic or biogenic alteration of kerogen, a fossilized organic material that is the source of oil and gas. In addition to providing the hydrocarbon source, kerogen plays a key role in developing reservoir quality in shales. Its degeneration creates pore space that makes up in part for the porosity lost during sedimentary compaction.
Because of its extremely high surface area and affinity for hydrocarbon molecules, this pore space is an excellent storage medium for gas, which becomes adsorbed onto the organic surfaces. In addition, free gas or oil may exist in larger pores, both within kerogen and between mineral grains. Understanding the mix between adsorbed and free hydrocarbons is essential for calculating total hydrocarbon content. Because of the role of kerogen in creating pore space and providing hydrocarbon storage, there is a strong correlation between kerogen content and total porosity, hydrocarbon saturation, and permeability. Therefore, kerogen content, or total organic carbon content (TOC), is an important indicator of overall reservoir quality.
Until now, TOC evaluation has involved indirect measurements and correlations. A new-generation spectroscopy measurement obtainable by using the Litho Scanner high-definition spectroscopy logging tool developed by Schlumberger provides a direct continuous measurement of carbon. Enabling the measurement are the tool’s new cerium-doped lanthanum bromide (LaBr3:Ce) gamma ray detector and advanced pulsed neutron generator. The tool also provides precise concentrations of 18 other elements—including most of the major rock-forming elements—which enable mineralogy to be determined. This is a key factor because when the carbon content in minerals such as calcite and dolomite is subtracted from total carbon, what remains is TOC. With mineralogy and TOC established, the determination of porosity is facilitated and adsorbed gas content can be estimated.
The remainder of the puzzle involves calculating the total hydrocarbon saturation. This is typically estimated from resistivity measurements, the interpretation of which can be uncertain in shales because of low porosity, high clay content, and unknown water salinity. However, the multifrequency dielectric dispersion measurement obtainable with the Schlumberger Dielectric Scanner tool can determine water volume independent of resistivity.
A third advanced measurement can be made by using a new application of the modular formation dynamics tester tool. From advanced interpretation of the pressure falloff following a stress test, it is possible to obtain the permeability and reservoir pressure. This information is not obtainable by conventional methods in reservoirs with very low permeability. Fig. 3 displays the results of an integrated reservoir quality evaluation. This plot summarizes all of the relevant properties that affect reservoir quality, including porosity, permeability, and fluid saturations, which include adsorbed and free gas content.
Once the reservoir quality has been determined, the completion quality must be quantified. Defined as the effective creation of maximum surface area per unit of reservoir volume during fracturing, this potential derives from the formation’s mechanical properties such as near- and far-field stresses and rock strength analysis. An integrated mechanical earth model (MEM) is created from drilling, log, and core measurements. The latest generation sonic log measurements, delivered by an advanced acoustic scanning tool, provide critical input to the MEM. By measuring axial, radial, and azimuthal slownesses in the formations, the tool can provide the vertical and horizontal Young’s modulus and Poisson’s ratio figures needed for the layered nature of the shale—which is calibrated to multistress tests of the core samples. In addition, the tool helps identify natural fractures, which can also be imaged precisely by using formation microimager tools.
Stress analysis is a critical input to the MEM. The tectonic stresses that control the stress anisotropy can be evaluated by analyzing drilling-induced fractures, as well as observations of shear wave splitting and crossover from dipole dispersion analysis. Additionally, discrete closure stress measurements performed by modular formation dynamics testers can confirm the stress profile and model from advanced acoustics, core analysis, and borehole image data. The presence and orientation of natural fractures also provide valuable information for designing hydraulic fractures. Including the natural fracture swarms with those of the hydraulic fractures, maximum reservoir contact can be obtained.
By combining the results of the reservoir quality step and the completion quality step, a “truth table” can be developed on which the selection decisions for the most promising completion zones are based. Data matches between the reservoir quality and completion quality results enable identification of the best hydraulic fracturing targets. The value of this approach is immense. Fig. 4 demonstrates the integration of reservoir quality and completion quality for the identification of good and bad treatment zones and optimization of treatment stages.
Using selective core analysis on the zones identified by a heterogeneous rock analysis technique, it is possible to acquire valuable information on reservoir quality and completion quality. TOC calculations can be verified, as can porosity estimates. Total and clay-bound water saturations can be confirmed, as well as matrix permeability, rock texture, and mineralogy. Other valuable core-derived data include those obtained from multistress tri-axial compression testing, proppant embedment studies, and fluid compatibility studies. If conventional core data has gaps or is missing altogether, large-diameter cores from a newly developed large-volume rotary sidewall coring tool can be used.
Recently, an operator in the Marcellus Shale play discovered that two of 11 fracture stages completed were contributing 70% of a well’s production. Evidence pointed to the arbitrary way the completion stages were placed at even intervals across the completion. By redesigning subsequent treatments according to the match between reservoir quality and completion quality data, the targets with the highest production potential were identified. An offset well in which completion was designed using stress mapping delivered the same production rate as the original well, but with a 50% completion cost reduction because of selective perforating. During previous treatments, the operator experienced screenouts on several stages and sustained costs of USD 300,000 each to mitigate the screenouts. Since switching to the new workflow, the operator has experienced no screenouts.
By using a systematic workflow and comprehensive formation evaluation, operators in shale plays can improve completion effectiveness, reduce completion costs, and boost production rates, well productivity, and field profitability.