Recovered energy generation using an organic Rankine cycle system.
There are more than 1200 natural gas compressor stations in the U.S. interstate natural gas pipeline network. The compressor stations maintain the flow of natural gas across the country (Figure 1). The very largest stations, designated by the larger circles on this figure, can move up to 4.6 billion [ft.sup.3] (130 million m3) of natural gas per day (U.S. DOE 2007).
[FIGURE 1 OMITTED]
Almost all of the pipeline compressor stations are fueled by a small fraction of the gas flowing through the pipeline. Although the older and smaller compressor stations may be powered by reciprocating engines or electric compressors, most of the 300 larger-scale stations are equipped with combustion turbines to drive centrifugal compressors. These larger stations represent more than 57% of the installed capacity (US DOE 2007). At almost all of these locations, the high-temperature turbine exhaust gas stream, still containing about 70% of the energy from the combustion process, is discharged to the environment. It is very difficult to find customers for this energy in the form of heat because most of these larger pumping stations are in very remote locations. However, electricity can be economically transported over long distances. If the turbine exhaust waste heat could be converted to electricity, significant improvements in overall energy efficiency would be achievable.
As long ago as 1979, studies were made to investigate technology capable of transforming this wasted heat into electricity (General Electric 1979). With the increase in fuel and electricity prices, the economic case for harvesting this wasted energy has become even more compelling. A Rankine cycle power system is often used to transform thermal energy into electrical energy. The most familiar Rankine systems include these four steps: (1) use thermal energy (in a boiler) to turn water into steam; (2) send the steam through a turbine, which in turn drives an electric generator; (3) condense the steam back into water by discharging the remaining thermal energy in the steam to the environment; and (4) pump the water back to the boiler. Such steam Rankine systems, powered by either coal or nuclear sources, provide most of the electricity generated in the United States.
Natural gas turbines are also used to generate electricity. While many gas turbine-generators operate as simple Brayton cycle systems, they have the potential for higher generating efficiency by combining the Brayton cycle with a Rankine cycle. In a combined-cycle system, the steam systems are powered by the exhaust from large gas-fired combustion turbines.
It is thus natural to consider steam Rankine systems as a possible method to recover the waste heat from the compressor station combustion turbines. Four bottoming steam systems were installed at pipeline compressor stations between 1968 and 1970 (General Electric 1979). Another such system was installed in the early 1980s near San Francisco (Tateosian and Roland 1983). However, by 2006, there were only three Rankine heat recovery systems in use at natural gas compressor stations. One of these three was initially constructed to work with a water/steam cycle, but freezing problems in Alberta, Canada, led to the installation of a system using an organic working fluid in 1999 (FERC 2006).
The organic Rankine cycle (ORC) is not new. Ships using acetone as motive fluid in piston engines were in service for some time between Europe and the Amazons in the late nineteenth century. A small solar turbine was operating in Libya in the 1930s. A clear theoretical analysis leading to criteria for fluid selection in a modified Rankine cycle was developed between 1958 and 1985, and related technology has been commercialized since 1965 (Bronicki and Schochet 2005; Tabor and Bronicki 1961; Tabor and Bronicki 1962; Bronicki 1972; Bronicki 1981; Bronicki 1988). As of March 2008, 2500 recovered energy generation (REG) systems for geothermal, solar, and heat recovery were in operation.
In an ORC, another working fluid, typically one chosen because it has a lower freezing temperature and other desirable properties, goes through the same four-step Rankine cycle process as a steam system. In the REG, the use of a recuperator and an intermediate heat transfer fluid widens the choice of the working fluid to optimize the heat-to-power efficiency and provides additional safety for the system operations. This intermediate fluid (thermal oil) is used in the waste heat oil heater (WHOH) to capture the waste heat from the turbine exhaust gas. The thermal oil selection is based upon low operating pressures, stability, and a low freezing point. The hot thermal oil from the WHOH is fed into the vaporizer and preheater of the REG where its thermal energy is transferred to the Rankine cycle's working fluid. Many compounds such as chlorofluorocarbons, ammonia, and hydrocarbons can be used to match the Rankine cycle to the level of heat available. Pentane was selected for this recuperated cycle and offers a good match for many industrial waste heat streams due to its thermodynamic properties and limited environmental impact.
The objective of the market transformation project described in this paper was to demonstrate the technical and economic feasibility of capturing thermal energy from a gas turbine driving a natural gas pipeline compressor by using a REG system based on a modified ORC that produces electricity with no additional fuel and near-zero emissions.
Four identical REG plants were designed, manufactured, and installed, and are owned and operated by subsidiaries of an ORC system manufacturer to provide base load power generation for an electric power cooperative utility (co-op). As part of the project, a co-op member built and owns the 69 kV transmission line and substation interconnection to the REG plants. The participating natural gas pipeline uses 13 simple-cycle gas-turbine-powered compression stations, each ranging from 35,000 to 40,000 hp (27 to 31 MW), to deliver natural gas to the upper Midwest region of the United States. Four of these stations, each with a 35,000 hp (27 MW) gas turbine, were selected by participating companies for REG. Each of the four compressor stations were retrofitted with REG plants with a design electricity generation capacity of 5.5 MW at each site. The compressor station near St. Anthony, ND, was evaluated in this project.
The business structure involves an alliance of the pipeline, the REG plant owner, and the co-op. The pipeline sells the resource (waste heat from the compressor station exhaust) to the REG plant owner. The REG plant is based on ORC technology and related auxiliaries developed by the manufacturer over four decades. The REG plant uses that resource to produce electricity and delivers power to the grid through a grid interconnection. The co-op purchases all of the power generated at the REG plant at a negotiated price under a 25-year Power Purchase Agreement. The REG plant was monitored for a period of one year to evaluate its performance.
Recovered Energy Generation Plant Design
The natural gas pipeline compressor station consists of a simple-cycle gas turbine driving a natural gas compressor to boost the pipeline pressure. The exhaust temperature from the gas turbine is roughly 900[degrees]F (480[degrees]C). A heat source at this temperature could be used for steam generation and is suitable for ORC plants. Steam technology used to recover residual heat from gas turbines in combined cycle electric utility plants can be cost-effective under the right conditions. However, gas turbines installed in compressor stations are about an order of magnitude smaller than their utility counterparts, increasing the installed cost per MW significantly, and are unattended and remotely controlled. Sufficient water supply is also a challenge for steam plants. Compressor stations are often installed in remote areas where in-situ water sources normally do not exist. In addition, high-pressure steam processes require licensed operators. Finally, as described previously, managing water at a remote site in a very cold climate can be difficult and expensive. For these reasons, an ORC system is the preferred solution for a REG for this type of project.
The ORC uses the thermal energy of the oil provided by the WHOH, which recovers heat from the gas turbine exhaust using thermal oil as the heat transfer fluid. The thermal oil provides energy to preheat and vaporize the hydrocarbon organic working fluid in the REG preheater and vaporizer heat exchangers. The working fluid (pentane) expands through a turbine directly coupled to a generator, then flows through a recuperator and condenses in an ambient air-cooled condenser. This process requires a certain amount of auxiliary electricity to operate fans and pumps, but produces more than six times the amount consumed by the REG plant. The following six processes in the ORC system (Figure 2) each change the working fluid state:
[FIGURE 2 OMITTED]
* Process 1-2: First, the working fluid is pumped from low to high pressure.
* Process 2-3: The working fluid is then heated in the recuperator and pre-heater.
* Process 3-4: The high-pressure liquid enters a vaporizer where it is heated at constant pressure by hot oil to become a saturated vapor.
* Process 4-5: The saturated vapor expands through a turbine to generate power output (ideally an isentropic process). This decreases the temperature and pressure of the vapor.
* Process 5-6: The vapor leaving the turbine enters a recuperator where it exchanges heat with the condensed working fluid discharged from the working fluid cycle pump.
* Process 6-1: The vapor then enters a condenser where it is cooled at constant pressure to become a saturated liquid. This liquid then re-enters the pump and the cycle repeats.
The manufacturer selects an indirect method of energy recovery from the gas turbine exhaust rather than direct heating of the working fluid (Figure 3) through the application of a proven thermal oil cycle. The use of the thermal oil cycle results in lower operating pressure at the WHOH and simplifies the system control by separating the gas turbine exhaust from the ORC. The completed REG plant at the pipeline compressor station (Figure 4) is intended to operate continuously throughout the year, except for periodic scheduled maintenance. Further design details are available in the project technical report (Sweetser and Leslie 2007).
[FIGURE 3 OMITTED]
[FIGURE 4 OMITTED]
Sequence of Operation
The sequence of operation is relatively straightforward. Gas-turbine exhaust heat is transferred to thermal oil circulating through the WHOH. The hot thermal oil evaporates the ORC pentane working fluid in the vaporizer and then gives up additional heat to the working fluid in the preheater before returning to the WHOH. Vaporized working fluid expands through the turbine and flows to the recuperator where it provides heat to partially preheat the condensed working fluid returning from the air-cooled condenser. A storage/ expansion tank accommodates any variations in thermal oil volume and maintains a constant head on the system. A gas turbine exhaust bypass stack and oil pump flow controller balance the WHOH conditions to avoid excessive cooling of the gas turbine exhaust gas and to control the maximum thermal oil temperature.
Data Acquisition System
The performance of the REG system was measured for a period of one year using the manufacturer's data acquisition system. Temperature and mass flow data were collected hourly to calculate desired performance parameters. Table 1 lists measured and calculated parameters. Table 2 lists the key heat input and power output state points used to determine peak and seasonal REG system energy efficiency (Figure 5).
[FIGURE 5 OMITTED]
Table 1. Measured and Calculated Performance Parameters in the Data Acquisition System Parameter Unit of Measure Date/hour MW Net power delivered to grid [degrees]F Turbine exhaust gas pressure inches water column Oil flow to WHOH lb/sec Exhaust gas temperature at WHOH inlet [degrees]F Calculated turbine exhaust temperature drop [degrees]F Oil temperature at WHOH outlet [degrees]F Oil temperature at WHOH inlet [degrees]F Calculated oil temperature rise [degrees]F Oil heat capacity at WHOH outlet Btu/lb. [degrees]F Oil heat capacity at WHOH inlet Btu/lb. [degrees]F Calculated heat transferred to oil Btu/h Calculated REG plant efficiency (heat to power) % Oil temperature at working fluid preheater inlet [degrees]F Oil temperature at working fluid preheater outlet [degrees]F Calculated oil temperature drop [degrees]F Working fluid temperature at condenser outlet [degrees]F Working fluid temperature at vaporizer outlet [degrees]F Table 2. State Point Measurement Locations State Point Measurement 1 WHOH entering gas temperature 2 WHOH leaving exhaust gas temperature 3 WHOH entering oil temperature and flow 4 WHOH leaving oil temperature 5 Electric power delivered to the grid 6 Working fluid vapor temperature 7 Condensed working fluid temperature
The REG plant was a first-of-its-kind project on a pipeline compressor station in the United States. It was expected that there would be technical challenges during this project that would reduce system availability during the initial operating period. The first year of the REG system operation included three phases, each of which had a different impact on overall system availability and delivered power (Figure 6). During the three month startup and commissioning period, there were several transition issues that reduced system availability. Most issues were related to controls and instrumentation adjustments. In addition, the pipeline compressor was occasionally shut down due to pipeline demand fluctuations not related to the REG plant. Issues experienced during the commissioning period were managed and controls were optimized without undue difficulty.
[FIGURE 6 OMITTED]
During the next four months, availability suffered from several components and controls malfunctions and failures, some of which were related to extreme low temperature freezing problems with sensors and valves. Faulty components and sensors were subsequently replaced with, for example, components with lower ambient temperature compatibility. Pentane pumps were upgraded to solve a mechanical problem that was discovered during commissioning. Availability was not affected, but peak output was reduced by about 1/3 when only one of the two parallel working fluid pumps was running. Specific issues identified and resolved during the shakedown period included:
* oil circulating pump replacement
* flawed diverter control logic
* nitrogen leak through a valve
* surge tank pressure relief valve failure
* frozen flow transmitter
* heat source valve transmitter failure
* air compressor failure
* pentane pump replacement
For the last five months of the monitoring period, the REG plant was available nearly continuously, except for when the pipeline compressor was shut down due to demand fluctuations. System peak output during this period was also affected by the pentane pump change-out period that continued into April 2007.
The project goal was to provide 5.5 MW base load power to the grid with at least 90% availability. The goal was difficult to meet during the first year of operation due to periods of inconsistent operation related to factors described in Figure 7. Increased projected natural gas pipeline sales and an improved understanding of system operating constraints obtained from this project are expected to significantly improve long-term availability and performance. The REG plant is expected to continuously achieve or exceed the 5.5 MW delivered power goal in the future and should consistently exceed 90% availability whenever the pipeline compressor is operating.
[FIGURE 7 OMITTED]
Data during the test period provides information on delivered power when the system was in operation (Figure 8). The average delivered power during REG operation was 5.57 MW with a standard deviation of 0.95 MW. The average output shows that the system performed well with respect to the contracted performance goal of 5.5 MW. However, there was significant variation in the delivered power to the grid due to various factors described above, mainly due to variations in the gas turbine operating load. For energy efficiency calculations and an evaluation of system behavior relative to ambient temperature, a subset of the data that avoids confounding factors such as pipeline compressor shutdowns was selected for analysis.
[FIGURE 8 OMITTED]
System Efficiency and Economic Considerations
Performance and economic viability are directly related to the load factor of the compression station, and the overall energy efficiency and availability of the REG plant. In any installation using waste heat, economic viability has two major components: WHOH sizing and design, and heat to electricity conversion efficiency. The first component affects the overall capacity of the system for a given waste heat source. The second component measures the heat conversion efficiency of the ORC heat conversion system. Optimizing both components is critical for economic viability. Auxiliary power consumption, used to drive the thermal oil and working fluid pumps and the fans for the air-cooled condenser, is also a factor.
Waste Heat Oil Heater Sizing and Effectiveness
The sizing and effectiveness of the WHOH is a critical element in the overall system performance. To avoid condensation conditions within the waste heat gas stream, the maximum amount of heat that could be harvested from the waste heat stream is 90% or less. The proportion that can be harvested is stated relative to the ambient temperature, and therefore includes the enthalpy that would be released as the water vapor present in the exhaust gases condenses. This convention is used to maintain consistency with the fuel input characterized by the higher heating value (HHV). To maintain an adequate margin of safety and to reduce maintenance costs, a slightly lower design limit (e.g., 88%) may be appropriate.
Ideally, the heat exchanger effectiveness would be calculated by comparing the heat available within the exhaust gas stream to the heat absorbed within the thermal oil. However, the mass flow rate of the hot gas was not measured, so this calculation was not possible. In the absence of hot gas flow data, the amount of heat available from the turbine exhaust was estimated using data for times when there was no flow through the WHOH bypass stack. For this system, there were periods with peak oil flow and no bypass at approximately 30[degrees]F (-1[degrees]C) ambient temperatures, so data for operation at this temperature were used in the analysis.
State points #1 and #2 in Figure 5 are the exhaust gas stream from the gas turbine. Because the actual exhaust gas composition and mass flow were not measured, the energy content was estimated from the combustion turbine heat rate and delivered power. For the turbine installed at the compressor station, the power and heat rate at International Organization for Standardization (ISO) standard conditions (59[degrees]F, sea level) are 35,000 shaft hp (26,100 kW) and 9534 Btu/kWh lower heating value (LHV) (10,559 Btu/kWh HHV). Based on published data, it was assumed that turbine efficiency increased by about 0.3% at 30[degrees]F compared to ISO conditions, and its output increased by about 7% (Energy Nexus Group 2002). The resulting assumed turbine efficiency for the calculations was estimated at 32.6% higher heating value (HHV), and its output increased to about 27,900 kW. Assuming 2% jacket heat losses, the total heat available in the exhaust gas was 191 million Btu/h (56,000 kW) as shown here:
* Higher heating value of fuel at 30[degrees]F (-1[degrees]C) Ambient conditions = 291 million Btu/h (85,400 kW)
* Fuel to power = 95 million Btu/h (27,900 kW)
* Jacket losses = 5 million Btu/h (1500 kW)
* Total heat in exhaust gas = 191 million Btu/h (56,000 kW)
Based on measured flow rates at state points #3 and #4 in Figure 5 and the variable heat content of the thermal oil at inlet and outlet conditions (using thermodynamic properties obtained from the thermal oil manufacturer), the energy transferred into the oil at a measured REG plant output of 7.3 MW was 167 million Btu/h (48,900 kW). This yields an estimated rate of heat that was harvested from the waste heat stream, as defined above, of 87%. The calculated percentage of heat harvested is consistent with a tightly controlled system to minimize the risk of condensation on the heat exchanger while maximizing efficiency.
Heat exchanger performance will vary as both the exhaust gas and thermal oil flow rates are controlled by the use of WHOH bypass or oil pump modulation. The calculated rates varied when oil pump flow control and WHOH bypass were used to maintain safe WHOH operation across a wide range of ambient temperatures, with significantly lower rates of about 65% at extreme low ambient temperatures.
Heat to Electricity Conversion Efficiency
The ORC plant conversion efficiency was calculated on an hourly and seasonal basis using mass flow and temperature data collected in the field coupled with manufacturer's rated data for the thermal oil. There was typically a large oil temperature rise in the WHOH (e.g., 177[degrees]F [81[degrees]C] at the inlet to 568[degrees]F [298[degrees]C] at the outlet). Since the specific heat of the thermal oil varies significantly as a function of temperature (e.g., 0.450 Btu/lb-[degrees]F [1.88 kj/kg-[degrees]C] at 200[degrees]F [93[degrees]C], and 0.589 Btu/lb-[degrees]F [2.47 kj/kg-[degrees]C] at 550[degrees]F [288[degrees]C]), the heat content calculation uses specific heat values at the WHOH inlet and outlet temperatures derived from the manufacturer's tabular data. Heat delivered to the thermal oil within the oil heater can be calculated from the manufacturer's data coupled with field data using the following equation:
[Q.sub.oil] = m([[c.sub.pout][T.sub.out]] - [[c.sub.pin][T.sub.in]]) (1)
[Q.sub.oil] = heat content in the thermal oil available to the ORC system
m = thermal oil flow rate
[c.sub.pout] = specific heat of oil at the WHOH outlet
[T.sub.out] = thermal oil temperature at the WHOH outlet
[c.sub.pin] = specific heat of oil at the WHOH inlet
[T.sub.in] = thermal oil temperature at the WHOH inlet
The conversion efficiency of the ORC system is calculated by dividing the net delivered electricity by the oil heat content as shown in the following equation:
[[eta].sub.ORC] = [([W.sub.turbine] - [W.sub.auxiliaries])/[Q.sub.oil]] (2)
[eta]ORC = ORC system conversion efficiency
[W.turbine] = total power generated by the REG turbine generator
[W.auxiliaries] = Auxiliary power required by the REG plant for controls, pumps, fans and other auxiliaries
During the yearlong data collection period, the natural gas pipeline flow--and compressor station waste heat availability--was highly variable. That and other confounding factors associated with commissioning and equipment issues rendered significant portions of the data inappropriate for an evaluation of the REG plant potential performance. Therefore, the dataset used for REG plant efficiency and peak output calculations was a subset of the complete test period data. Because the compressor operation itself was not monitored, and the impact of other factors was highly variable, proxy criteria were defined to delete those data observations that corresponded to low or no waste heat availability. The criteria used to indicate adequate waste heat availability were:
* hot oil supply temperature to the vaporizer greater than 515[degrees]F (268[degrees]C)
* minimum heat available in the thermal oil greater than 112 million Btu/h (32,800 kW)
Ambient temperature had a noticeable impact on overall REG plant efficiency and power output. The ORC system consistently performed between 13% and 15% thermal efficiency when the REG system was operating at full power during the test period, with peak efficiency occurring between 20[degrees]F and 50[degrees]F (-7[degrees]C and 10[degrees]C) ambient conditions (Figure 9).
[FIGURE 9 OMITTED]
Effect of Ambient Temperature on Recovered Energy Generation Plant Power
Ambient temperature had a significant effect on peak power delivered to the grid (Figure 10). Peak output increased from 6.5 MW at 0[degrees]F (-18[degrees]C) to 7.1 MW at 32[degrees]F (0[degrees]C), falling to 5.3 MW at 95[degrees]F (35[degrees]C). The following equation provides a polynomial curve fit of the data used in this analysis:
[FIGURE 10 OMITTED]
[W.sub.grid] = 6.48 + 0.035[T.sub.ambient] - 0.0005[([T.sub.ambient]).sup.2] (in[degrees]F)
[W.sub.grid] = 7.09 + 0.054[T.sub.ambient] - 0.0016[([T.sub.ambient]).sup.2] (in[degrees]C) (3)
[W.sub.grid] = net power delivered to the grid, MW
[T.sub.ambient] = ambient temperature
Measured system performance data and calculated ORC efficiency values were also used to examine major constraints that impact REG plant output at different ambient temperatures. Gas turbine exhaust temperatures, thermal oil temperatures, condensed working fluid temperature, and peak power were tracked relative to ambient temperature (Figure 11). As expected, the power output dropped at higher ambient temperatures. The gas turbine exhaust temperature increased at higher ambient temperatures, which normally would translate into increased power output. However, stack bypass was used to avoid overheating the thermal oil at warm ambient temperatures. This limited the waste heat recovery capacity and reduced overall output and efficiency at higher ambient temperatures.
[FIGURE 11 OMITTED]
To reduce the REG plant capital cost, the WHOH and exhaust stack were made from carbon steel. The exhaust gas temperature at the WHOH outlet was controlled using oil flow modulation, especially at cold ambient temperatures, to avoid condensing water vapor and associated corrosion. This limited output and efficiency at cold ambient temperatures.
To further understand the impact of different ambient conditions on system performance, the data set was split into three performance regions:
* low temperature (below 25[degrees]F [-4[degrees]C])
* peak performance temperatures (between 25[degrees]F and 50[degrees]F [-4[degrees]C and 10[degrees]C])
* high temperature (above 50[degrees]F [10[degrees]C])
Power output had a distinct negative change of slope as ambient temperature fell below 25[degrees]F (-4[degrees]C). The dominant factor in this reduction appears to be the reduction in absolute gas turbine exhaust temperature related to lower ambient temperature (Figure 11). The working fluid (pentane) condenser categorically improved performance at lower ambient temperatures. Pentane condensate temperatures fell from 100[degrees]F (38[degrees]C) at 80[degrees]F (27[degrees]C) ambient temperature to 34[degrees]F (1[degrees]C) at 0[degrees]F (-18[degrees]C) ambient temperature. The WHOH performance was a function of higher gas turbine exhaust gas mass flow and lower exhaust gas temperature, with an apparent negative slope change as the exhaust temperature went below about 875[degrees]F (468[degrees]C). The lower gas turbine exhaust gas temperature difference started to dominate the higher mass flow rate at ambient temperatures below 25[degrees]F (-4[degrees]C), which can be seen in the change of slope of the exhaust gas and thermal oil temperature differences in Figure 9. In this region, variable oil flow to the WHOH maintained an exhaust gas temperature above 200[degrees]F (93[degrees]C) to avoid condensation in the heat exchanger.
The system reached its peak performance at ambient temperatures between 25[degrees]F and 50[degrees]F (-4[degrees]C and 10[degrees]C). The balanced plant operation limited the need for WHOH bypass and thermal oil flow rate modulation in this region. The oil temperature was between 580[degrees]F and 600[degrees]F (304[degrees]C and 315[degrees]C) leaving the WHOH over much of this ambient temperature range, while the exhaust gas temperature was between 210[degrees]F and 225[degrees]F (99[degrees]C and 107[degrees]C).
Above 50[degrees]F (10[degrees]C) the drop-off in power was roughly a 370 kW/[degrees]F (205 kW/ [degrees]C) increase in ambient temperature. The major constraint in this region was the recommended maximum operating temperature of 625[degrees]F (329[degrees]C) for the thermal oil. This reduced the power output by both reducing the cycle efficiency and by reducing the amount of heat that could be harvested from the waste heat resource. The efficiency reduction was related to the condensed working fluid temperature, which increased as ambient temperature increased. The increased condensed working fluid temperature was caused by the air-cooled condenser capacity reduction as ambient temperature increased. Because of the upper limit on the thermal oil temperature, it was not possible to offset the increased condensing temperature with an increased vaporizing temperature, so the temperature difference was reduced by 9%. The upper limit on efficiency for all Rankine cycles is the Carnot efficiency, which is proportional to the difference between the vaporizing and condensing temperatures as shown in Equation 4. The reduction in the temperature difference therefore translates into a direct reduction in cycle efficiency of approximately 9%.
[[eta].sub.Carnot] = [([T.sub.vaporizing] - [T.sub.condensing])/[T.sub.vaporizing]], (4)
[[eta].sub.Carnot] = Carnot energy conversion efficiency,
[T.sub.waporizing] = temperature of the working fluid gas leaving the evaporizor, and
[T.aub.condensing] = Temperature of the working fluid liquid leaving the condenser.
The increased condensed working fluid temperature also reduced the WHOH's ability to capture heat from the gas turbine exhaust gas due to the oil temperature limit. Hot gas flow through the WHOH was controlled using the bypass stack as the exhaust gas temperature increased to avoid overheating the thermal oil as the working fluid temperature increased. The heat gain by the thermal oil at 50[degrees]F (10[degrees]C) was 160 million Btu/h (46,900 kW) and at 90[degrees]F (32[degrees]C) was reduced to 145 million Btu/h (42,500 kW), a 9% reduction. Although auxiliary power consumption was not measured directly, the system is designed to vary the fan power used for the air-cooled condenser. The highest level of this fan power would correspond to the higher ambient temperatures. These three factors (reduced cycle efficiency, reduced heat input, and increased fan power) combined to reduce the net REG power output by about 20% (from 7.0 to 5.6 MW) at the highest ambient temperatures.
It is anticipated that approximately 43 million kWh will be supplied to the grid per year, with projected annual operating and maintenance costs of approximately $200,000 per year. The capital cost for the 5.5 MW REG system was approximately $2,500 / kW. It is estimated that future projects would require a minimum electricity purchase price of approximately $0.05/kWh, based on projected output for an acceptable return on investment.
Various contract terms (15 to 25 years in duration) and cost of capital ranging from 6% to 10% for clean energy projects were considered to determine the net present value (NPV) of this project. Positive NPV values ranged from $2 million to $12 million. The internal rate of return ranged from a low of 5% for a 15-year contract to a high of 15% for a 25-year contract. These values did not include any federal or state subsidies for providing pollution-free electricity.
Based on the successful results of this project, the utility company plans to purchase power from four new REG plants that the manufacturer plans to construct, own, and operate along a natural gas pipeline in Montana, North Dakota, and Minnesota. The new REG power plants are in addition to the existing four facilities in North and South Dakota that have been in commercial operation since fall 2006. Twenty-two additional MW of electricity from the four new REG power plants will be sold to the utility under a long-term power purchase agreement that is expected to add up to approximately $6.4 million in yearly revenues for the REG plant owner. Eleven MW of power from an additional two sites will be contracted to two other utility companies. Development of the additional six sites will increase the capacity of the owner's portfolio of REG plants to approximately 55 MW.
SUMMARY OF LESSONS LEARNED
Lessons learned that may be useful for evaluating future REG plant opportunities include:
* When operating at full recovered heat input, the REG plant consistently delivered 5.5 MW or more output to the grid at up to 15% conversion efficiency (net electricity output/recovered heat input). The REG plant improved the overall energy efficiency by 28%, from 32% simple cycle efficiency to 41% for the combined system. The system entered its second year of operation with the expectation of consistently achieving near 100% availability when the pipeline compressor is operating.
* There is an economic model that makes existing ORC technology applied to pipeline compressor stations cost-competitive with other generation technologies.
* Remote pipeline-based REG systems can provide basel-oad power.
* Cold ambient operation provided challenges, including the need to replace frozen flow transmitters and change certain valve designs that were prone to freezing.
* There was minimal environmental impact, minimal permitting, and virtually zero incremental emissions related to the compressor station installation.
* The pipeline compressor was shut down several times during the test period due to market demand fluctuations. Since compressor downtime affects annual waste heat availability and baseload power output, it is important to obtain good estimates of annual compressor run hours from the pipeline when selecting project locations.
* The operating constraints identified were: heat input (temperature and mass flow) that was dictated by the gas turbine's operation, the limit on minimum WHOH outlet exhaust gas temperature to prevent condensation, and the maximum thermal oil operating temperature limit.
* An area for potential improvement in performance that should not negatively impact capital cost is to consider eliminating the thermal oil loop. Risks of alternative approaches to a thermal oil loop must be carefully analyzed before making any recommendation.
The project described in this paper was developed within the National Accounts Energy Alliance in response to a solicitation issued by the Oak Ridge National Laboratory on behalf of the U.S. Department of Energy (DOE). Under the resulting collaboration between the DOE, Gas Technology Institute, and Oak Ridge National Laboratory, a field research test and verification project was conducted at the Recovered Energy Generation System at Northern Border Pipeline Company Compressor Station #7 near St. Anthony, ND. REG system equipment was designed, manufactured, supplied, and installed by ORMAT Technologies, Inc. Basin Electric Power Cooperative is purchasing the electricity under a purchase power agreement with an ORMAT subsidiary, which owns and operates the plant.
Bronicki, L.Y. 1972. The Ormat Rankine power unit. IECEC, San Diego, CA.
Bronicki, L.Y. 1981. Energy recovery from waste heat by organic Rankine cycle turbogenerators. Institute of Marine Engineers, London.
Bronicki, L.Y. 1988. Experience with high speed organic Rankine cycle turbomachinery. Lappeeranta, Finland.
Bronicki, L.Y., and D.N. Schochet. 2005. Bottoming organic cycle for gas turbines. American Society of Mechanical Engineers, Reno, N V.
Energy Nexus Group. 2002. Technology characterization: gas turbines. Environmental Protection Agency, Climate Protection Partnership Division. Washington, DC.
FERC. 2006. Final environmental impact statement for the proposed rockies western phase project. Docket Nos. CP06-354-000, CP06-401-000, and CP06-423-000, Federal Energy Regulatory Commission, Washington, DC.
General Electric. 1979. Pipeline bottoming-cycle study: operational reliability and maintainability assessment, GESP-813, General Electric Company, Cincinnati.
Sweetser, R., and N. Leslie. 2007. Subcontractor report: National account energy alliance final report for the basin electric project at northern border pipeline company's compressor station #7, North Dakota, ORNL/ TM-2007/158, Oak Ridge National Laboratory, Oak Ridge, TN.
Tabor, H., and L.Y. Bronicki. 1961. Turbine for small solar power package. United Nations Conference on New Sources of Energy, Rome.
Tabor, H., and L.Y. Bronicki. 1962. U.S. Patent No. 3040528
Tateosian, C.J., and G.K. Roland. 1983. Installation of a waste heat recovery system at a gas turbine-driven compressor station. Proceedings of the American Society of Mechanical Engineers International Gas Turbine Conference, Phoenix, AZ.
U.S. Department of Energy. 2007. Natural gas compressor stations on the interstate pipeline network: developments since 1996. Energy Information Administration Office of Oil and Gas, Washington, D.C.
Neil P. Leslie, PE
Richard S. Sweetser
Therese K. Stovall, PE
Neil P. Leslie is a research manager in the End Use Solutions Sector at Gas Technology Institute, Des Plaines, IL. Richard S. Sweetser is the president of Exergy Partners Corporation, Herndon, VA. Ohad Zimron is the vice-president of U.S. operation at Ormat Technologies, Inc., Reno, N V. Therese K. Stovall is a senior research engineer in the Engineering and Science and Technology Division at Oak Ridge National Laboratory, Oak Ridge, TN.