CC BY
10
Solar towers with thermal storage for integrated solar
combined cycle plants
Juergen H. Peterseim,
PhD, University of Technology in Sydney, Australia Dr. Amir Tadros,
Group-Mechanical Engineering leader, Laing O'Rourke, Australia Prof. Udo Hellwig,
ERK Eckrohrkessel GmbH, Berlin, Germany Prof. Stuart White,
Institute for Sustainable Futures at the University of Technology, Sydney, Australia
Hybridization is a low-cost approach to implement power plants worldwide. Generally, integrated solar combined cycle power stations combine the energy harvested from solar radiation with another fuel to cut fuel costs and environmental impact. Several operating plants would qualify for conversion into integrated solar combined cycle ones, and plants in the planning stage should consider this option. The example described below investigates the conversion of a Western Australia power plant into an integrated solar combined cycle plant using a solar tower and thermal storage.
Keywords: power plant, combined cycle, hybridization, renewable energy, solar tower.
Current global efforts to reduce CO2 emissions aim to incentivize utility and industrial plant operators to build renewable and low-carbon intensity generation assets. With operators looking for the most cost effective power generation possibility, the use of renewable energy sources in existing assets or new hybrid plants is appealing. Renewable energy technologies, such as concentrated solar power (CSP), progressed significantly in the past decade in regard to efficiency and cost. However, when making investment decisions without including external costs for fossil fuels or government incentives for renewable electricity, the current economic climate favors natural gas-fired open cycle gas turbines (OCGT) and combined cycle gas turbines (CCGT) plants when compared to concentrated solar power installations.
To lower the specific investment of CSP systems and prepare for rising natural gas prices in the future, existing OCGT and CCGT plants could be retrofitted and new plants designed with a CSP component. This would create environmental benefits, such as carbon intensity reduction, as well as economic benefits, such as lower investment CSP installations and renewable energy certificates.
Of particular interest is the retrofit of solar towers to reduce cost and allow independent operation of CCGT and CSP component. Current ISCC plants use parabolic trough systems with thermal oil to boost the steam output of the CCGT's heat recovery steam generator (HRSG). This is a well-established concept, but limits the CSP contribution and requires both plants to operate simultaneously.
Depending on the load profile, the integrated solar combined cycle plant could operate with and without
storage. Thermal storage would be valuable to provide solar generated electricity during peak demand times in the evening, but the currently high costs for thermal storage of 90 $/kWhth makes it hard to justify this investment without a premium on power dispatchabi-lity.
Typically, daytime electricity prices are significantly higher than nighttime prices. This fits well with integrated solar combined cycle (ISCC) concepts, since they provide additional solar generated electricity during the day and evening period, which helps recover the higher specific CSP investment.
ISCC requirements
ISCC plants seem to be a very suitable technology to reduce the carbon intensity. Natural gas prices are expected to increase significantly, with new liquefied natural gas facilities coming online. Additionally, the CSP component will benefit from current and expected increases in electricity prices.
From an economic perspective, ideal locations for ISCC facilities are remote load centers since electricity prices are higher there than along the east coast, where a grid distributes large quantities of low-cost electricity from brown and black coal-fired power plants. Particularly interesting are areas with existing OCGT or CCGT plants, since conversion costs would be comparatively low.
Several operating plants would qualify for ISCC conversion, and plants in the planning stage should investigate the ISCC option for its economic viability in the feasibility study stage. If the DNI is sufficiently high, the developer should at least reserve certain areas for a future plant conversion.
Typically, an average daily DNI of >20 MJ/m2 is required for a stand-alone CSP plant, but due to cost reduction benefits of ISCC plants through the joint use of plant equipment, such as steam turbine and condenser, areas with a DNI >17 MJ/m2 could be considered as well.
Current ISCC concepts
Currently, a number of ISCC plants are in operation around the world, such as the U.S., Egypt and Morocco. The largest ISCC unit in operation is the 75 MWe equivalent Martin Next Generation plant in Florida [1]. All existing ISCC plants use the mature parabolic trough technology with thermal oil, providing saturated or slightly superheated steam to the heat recovery steam generator (HRSG). In the HRSG the steam is further superheated, using the hot gas turbine exhaust to meet the steam turbine requirements. This concept is well proven in several ISCC plants but requires both a CCGT and CSP plant to operate simultaneously. Typically, the CSP contribution is below 15 percent of the plant capacity.
Due to the use of parabolic trough with thermal oil, CSP steam temperatures are limited to <390 0C since the thermal oil degrades very quickly above 400 0C. To increase the steam enthalpy provided to the host plant, other working fluids such as water-steam or molten salts have to be used. One 5 MWe prototype parabolic trough plant using molten salt and thermal storage was retrofitted to the Priolo Gargallo CCGT plant in Italy, and CSP steam temperatures reached 535 0C. This is a benefit compared to the parabolic troughs with thermal oil, but with significantly larger solar tower plants in operation or beginning operation soon, it is expected that such plants are easier to finance. Additionally, molten salt solidification is of significantly less concern in a solar tower compared to parabolic trough system.
Solar tower ISCC
With more solar tower plants commencing operation and construction around the world, such as Gemasolar in Spain and the Ivanpah and Crescent Dunes projects in the U.S., the technology is becoming increasingly mature and bankable. Significantly higher steam qualities (>540 0C and >110 bar), compared to parabolic troughs with thermal oil, allow the simple and more efficient integration into the CCGT's high temperature/pressure steam cycle. An air-cooled condenser is chosen to minimize water consumption.
The benefits of all ISCC plants are cost savings through the joint use of equipment, such as the steam turbine, condenser and feedwater system. However, ISCC plants using a solar tower have the additional benefit of sharing building infrastructure, e.g. the main stack could be modified to support the solar receiver. Typically, only utility-scale CCGT plants have stack heights >100 m, but raising the lower stack height of an industrial HRSG, for example 70 m, is technically possible and provides an ISCC cost saving opportunity as, according to a 2011 study by Hinkley et al, the tower in
a stand-alone plant requires around 5 percent of the total investment.
Obviously, not the full 5 percent could be saved, but a significant portion. The conversion of OCGT to ISCC plants not only increases the natural gas conversion efficiency, but also reduces the investment required as a significant portion of the equipment is already installed, e.g. the gas turbines, gas supply infrastructure, roads and control room.
Thermal storage references exist for parabolic trough, up to 7.5 h full-load operation, and solar tower plants, up to 15 h full-load operation. Implementing thermal storage has the potential to maximize the CSP contribution, but strongly depends on the financial value of energy dispatchability.
Historically, electricity prices are higher during the daytime and that is when the CSP plant can provide the additional capacity needed. With energy demand remaining high into the evening, a 3 h molten salt thermal storage is considered in this concept. This allows additional power generation, but limits the investment in the capital intensive technology. The CCGT component can provide electricity during the night at a lower cost.
Port Hedland case study, Western Australia
The following example investigates the conversion of the 110 MWe (at 25 oq OCGT power plant at Port Hedland, Western Australia [2], into a 200 MWe gross (194 MWe net) ISCC plant using a solar tower with molten salts and 3 h of thermal storage. Port Hedland is a suitable site for a CSP facility with a DNI >2,400 kWh/m2/year, and higher than average gas and electricity prices. However, it is a cyclone affected area, so all equipment has to be designed accordingly.
The technical and economic data presented in this case study derive from modeling work with Thermoflex version 23.0. The investment data derived from the software package was adapted to accommodate higher installation prices in remote areas.
Port Hedland Power Station is a 110 MW open cycle gas-fired power station located in the Pilbara region of Western Australia. The OCGT plant consists of three 36.7 MWe (at 25 oc and 40 MW peak rated) gas-fired turbines connected to the North West Interconnected System by a 66 kV transmission line. The plant was installed in 1996.
ISCC conversion
The OCGT to ISCC conversion would increase the peak electricity output by 90 MW, a 60 MWe equivalent from HRSG and 30 MWe equivalent from CSP. The gross electric efficiency of the CCGT part would be 46.4 percent, assuming temperature of 25 0C. This is a very efficient use of natural gas compared to back-up boilers in traditional CSP plants, <28 percent. However, it is lower than efficiencies of other CCGT facilities in operation worldwide, up to 60 percent, as the high daytime ambient temperature, 260 days/year
№2 (68) 2016, MapT-aipem,
y. ulartia mcj actm a m a h m ¡ i i i hh irn ¡ i m a wsi! a g m w m i i b ¡itwia
>30 0C and 135 days/year >35 0C, affects the performance of the gas turbines and air-cooled condenser.
The double pressure HRSG generates a maximum of 143 t/h at 515 0C and 120 bar and 53 t/h at 15.5 bar. The solar tower capacity is 85 t/h at 515 0C and 120 bar. Steam reheating is not typical at this size HRSG, but considered in this study to maximize the efficiency of the capital intensive CSP component. Reheating to 515 0C (35 bar) occurs in the HRSG and solar tower separately. Reheat steam turbines are available at this capacity from several vendors.
This allows a generation capacity of up to 200 MWe during CSP operation and 170 MWe with only the CCGT. The solar tower would require a height of 120 m and the heliostat field is designed with a solar multiple of 1.7 to optimize the CSP contribution.
During CCGT operation, an integral deaerator in the HRSG and bleed turbine steam preheats the feedwater, while in CSP-only operation a conventional feedwater heater would compensate the integral deaerator shortfall. The conventional feedwater heater system is required to allow CSP only operation.
This study uses air-cooled condensers rather than a watercooled system, since water is a scarce and expensive resource in the region. When converting the current OCGT to CCGT, plant capacity would increase to 164 MWe (net), but during high ambient temperatures - sometimes up to 49 0C occurs in this region -the CCGT output would decrease at times where electricity demand increases. The ambient temperature effects on the power station from 20 0C to 45 0C. A CSP component could compensate this shortfall and keep electri-city output high until 9 p.m., after which time demand typically decreases. Adapting the plant output to electricity demand is best done with the GT's or the heliostat field until the thermal storage is completely charged.
One option to keep the plant output higher is the use of low-temperature CSP heat to chill the GT's inlet air. Gas turbine inlet air cooling has been realized, but not yet with CSP. However, this concept is not modeled in this paper and is an area for future investigation.
Plant layout
With the Port Hedland power station being a remote facility, there are only a few constraints for locating new equipment. These include the location of the gas turbines, the switchyard and the power lines. It is assumed that none of the existing infrastructure can be moved. Therefore the heliostat field and the HRSG are arranged south of the current OCGT plant. The HRSG's stack is being used to support the solar receiver at a height of 120 m. Structural modifications are required to ensure stability.
The exhaust from all three gas turbines would be combined in a single gas duct and supplied to one vertical HRSG. The use of one HRSG per GT is more common, but this solution increases investment. Also, the three GTs will operate simultaneously to minimize differences in gas flows. The single HRSG would be loca-
ted along a north-south axis on the Port Hedland facility with the stack/solar tower being farthest south. The molten salt tanks are arranged a few meters south of the HRSG, with the molten salt/water-steam heat exchangers between them.
The heliostat field would be arranged around the tower in a 130-degree circle and covers an area of 1.4 km2. A full circle would reduce the distance between receiver and furthest heliostats, lowering optical losses, but it is not possible due to existing infrastructure, particularly outgoing transmission lines.
Steam turbine and air-cooled condensers are arranged close to the HRSG to minimize piping and heat losses. Existing buildings could be upgraded to accommodate the additional control and auxiliary equipment. The HRSG is expected to be an outdoor installation with only a roof to protect valves, etc., from severe weather effects. The steam turbine would be housed in a designated building.
Economic viability
The investment required to convert the current 110 We OCGT plant into a 200 MWe ISCC facility is expected to be around AUD 440 m, including 9 percent owner's soft costs for permits, finance, project management, etc.
The calculation is based on plant commissioning in 2016, a 30-year plant lifetime, 8 percent interest rate, 7 percent discount rate, as well as 6,000 h full-load CCGT and 2,500 full-load CSP operation as well. The modeling includes a price on carbon emissions of 23 AUD/t and escalation rates for electricity, natural gas and water.
Depending on gas and electricity prices, the payback for the new facility would range from 9-25 years. This is 15 percent shorter than the payback for a 200 MWe greenfield ISCC plant since GT equipment and fuel infrastructure is already in place. To make the project viable, an average equalized cost of electricity of > 120 AUD/MWh is required. Due to the age of the existing GT's, it is expected that these would need replacement or a major refurbishment during the operational life of the new ISCC facility.
Expected solar tower investment reduction of 28 percent by 2020 would increase the internal rate of return of scenario 3 percent to 10.8 percent and scenario 4 percent to 13.9 percent. However, these cost reductions require the ongoing constructions of such plants to realize learning curve benefits.
The use of the solar tower technology is promising to enhance the CSP contribution of ISCC plants since it enables identical steam parameters to the HRSG with independent operation of the CCGT and CSP components. This is significantly different to current ISCC plants where the CSP component provides significantly lower temperature steam to the HRSG for additional superheating. Also, the joint use of not only plant equipment, such as steam turbine and condenser, but also plant infrastructure, stack to support solar receiver, has the potential to lower the capital requirements
for ISCC installations. The conversion of existing OCGT into an ISCC plant is very promising since it results in a 15 percent shorter payback period compared to building a greenfield ISCC plant. A critical part of the implementation of such ISCC systems is the continuous successful operation of existing solar and new large
scale solar tower installations, as currently under construction in the U.S. With current solar tower projects mainly limited to activities in the U.S., hybridization would be a lower cost approach to implement further plants worldwide and advance the technology along the learning curve.
References
1. Martin Next Generation Solar Energy Center. Available at: www.cspworld.org/cspworldmap/martin-next-generation-solar-energy-center (accessed 16 February 2016).
2. OCGT power plant at Port Hedland, Western Australia. Available at: www.globalenergyobservatory.org/geoid/43476 (accessed 16 February 2016).
№2 (68) 2016, MapT-anpem
