footprint through increased overall cycle efficiency

Transcription

1 Reducing the CO 2 footprint through increased overall cycle efficiency By F. Cinelli, A. Miliani, G. seghi / GE Oil & Gas, M. Lehar / GE Global Research Center GE imagination at work

2 GE Oil & Gas technology insights Reducing the CO 2 footprint through increased overall cycle efficiency By F. Cinelli, A. Miliani, G. seghi / GE Oil & Gas, M. Lehar / GE Global Research Center Abstract With the growing need to address greenhouse gas emissions globally, significant effort is being undertaken to identify the most promising technical options for reducing CO 2 emissions resulting from human activity. This paper provides an overview of a technique to reduce CO 2 emissions from fossil fueled plants using a specific technology to increase efficiency. Improvements in cycle efficiency can substantially contribute to reducing the CO 2 emissions per unit of useful energy produced. In this context, the recovery of low temperature heat using an Organic Rankine Cycles (ORC) is an attractive alternative to traditional steam cycle technology. Examples of related GE research and development are included in this paper to demonstrate attractive opportunities for reducing CO 2 emissions per MW of energy produced with ORC technology. Moreover, there are a number of GE upgrades targeted for gas turbine efficiency and emissions, which can provide a wide array of efficiency improvements and a significant reduction in CO 2 footprint. Introduction We are currently in a deep economic recession, but in spite of that, the long term projection shows that per capita spending is growing. In 195, 2.5B people spent $5T, $2k/Capita. The projection to 218 is that 5.5B people will spend $88T, $16k/Capita. This increase in economic activity means more energy is required to support the growth. In addition, we need to consider the other aspect of the long-term market environment: the end of easy oil. Existing oil and gas wells are being depleted at an average rate of 4.5%. To sustain the growth, new wells need to be discovered, or production from existing ones enhanced. Assuming a projected oil demand of 1.1% year over year, in 218 an additional 48 Mbbl/d will be required. These market realities make the community and governments aware of the need to use energy more efficiently. It pushes the community to set new rules for governments to avoid wasting energy, to stimulate the deployment of renewable energy technology, and to regulate carbon dioxide emissions to reduce the impact of human activities on the environment. Recovery of waste energy is a commitment that many governments are taking under pressure from their constituents. The combustion of hydrocarbons is the process enacted by humans with the highest impact on CO 2 emissions. CO 2 generation is directly tied to the efficiency of energy conversion from hydrocarbons. One the most effective approaches to increasing the efficiency of energy conversion plants is waste heat recovery. In this paper we will focus on waste heat recovery from gas turbine exhaust streams. Modern plants designed for base load power generation and equipped with gas turbines typically employ a bottoming cycle. The standard bottom cycle is the Rankine cycle. The gas turbine is the topping cycle while the water/steam Rankine cycle recovers the heat from the exhaust of the gas turbine as a bottoming cycle. The combination of the two is referred to as the combined This paper was presented at the GASTECH 29 in Abu Dhabi, UAE, May 25-28, 29. It was selected for presentation following a review of the abstract submitted by the authors. 1

3 cycle. This thermodynamic configuration allows modern plants to exceed 5% thermal efficiency, even though the gas turbine simple cycle ranges from 3 to 4%. The limitations of the water/steam Rankine cycle prevent its application as a waste heat recovery process for all gas turbine exhausts. The large thermal inertia of the mass of water circulating in the system and the limited operability range of the Rankine cycle means that it fits as a bottoming cycle when the gas turbine is operated at steady conditions or with a limited load variation. The other key roadblock to the spread of the combined cycle employing water/ steam is water consumption loss from the sealing and vacuum system means that a substantial water source must be available near the site. Currently more than 85% of the installed gas turbines used in mechanical drive applications are simple cycle. GE Oil & Gas simple cycle units account for a total output power of 2GW. This could potentially generate an additional 4GW through heat recovery. To give an idea of the potential significance of ORC bottoming cycles, if all these simple cycle gas turbines were coupled with the GE ORC system (ORegen) it would be possible to produce enough power to satisfy the needs of more than 2 million homes in the UK. This paper introduces the Organic Rankin Cycle as an alternative waste heat recovery system in place of the standard water/steam system. GE ORegen is capable of efficiently recovering the waste heat from gas turbines operated in mechanical drive applications where the requirements of the plant imply an extended operability range between 5 and 1% of load and plant locations that cannot support any water makeup. In addition to introducing the ORegen system, a case study based on a pipeline compressor station using simple cycle gas turbines as drivers is presented. The case study includes the business implications of adopting ORegen as a waste heat recovery system. The ORC concept The Organic Rankine Cycle (ORC) is a thermodynamic cycle based on the Rankine Cycle principle in which the working fluid is an organic, high molecular weight fluid with a liquid-vapor phase change occurring at a lower temperature than that of water. The cycle permits the energy of the heat source to be converted into useful work, which can in turn be converted into electricity using a generator. Temperature [ºC] ORC cycle T-S digram Entropy [kj/(kgºc)] FIGURE 1 ORC T-S diagram T dew T bubble T min. p T max. p Heaters Expander Recuperator Condenser Pump Working principle of ORC The working principle of the Organic Rankine Cycle is the same as that of the Rankine Cycle. The working fluid is pumped into a heat exchanger where it is evaporated, passes through a turbine and is then finally re-condensed. In the ideal cycle, the expansion is isentropic and the evaporation and condensation processes are isobaric. In the real cycle, the presence of irreversible processes lowers the cycle efficiency. Those irreversibilities mainly occur: 1. During the expansion: Only a part of the energy recoverable from the pressure difference is transformed into useful work. The other part is converted into heat and is lost. The efficiency of the expander is defined by comparison with an isentropic expansion. 2. In the heat exchangers: The working fluid path ensures effective heat transfer but causes pressure drops that lower the amount of power recoverable from the cycle. 2

4 ORC application The Organic Rankine Cycle technology can potentially be used in several applications: Air cooled condenser vaporized cyclopentane liquid cyclopentane diathermal oil Waste heat recovery Waste heat recovery is the main application for ORC systems. The Organic Rankine Cycle is a suitable system for recovering waste heat from industrial and agricultural processes such as gas turbine exhaust gases, hot exhausts from cement plants, refinery processes, gas engines, fermentation of organic products, hot exhausts from furnaces, etc. Geothermal plants The ORC can be used to recover energy from geothermal heat sources. The cycle efficiency for low temperature geothermal sources (typically less than 1 C), is relatively low and depends strongly on the heat sink temperature (defined by the ambient temperature). Biomass power plants Biomass is available worldwide and can be used for the production of electricity in small to medium scale power plants. Solar thermal power The Organic Rankine Cycle can be used with solar parabolic trough technology in place of a typical steam Rankine cycle. GE s research on solutions based on the ORC concept have focused mainly on the waste heat recovery application and in particular: warm water Oil 6 kw < 9ºC Jacket pump Engine jacket 19 kw 4 bar < 82ºC LT loops < 94ºC HT intercooler 49 kw Preheater (brazed plate) Pump 2,7769 kg/s 3 bar Inlet mixture 2 bar X = HT loops FIGURE 2 Gas engine ORC heat recovery system 18ºC Pump 1 Pump 3 Cyclohexane Condenser (air blown finned tubes) 15ºC X = FIGURE 3 GT ORC waste heat recovery system GE Oil & Gas and GE Global Research have jointly developed ORegen as a commercial solution for GT waste heat recovery applications. ORegen ORegen is a thermodynamic superheat cycle that recovers waste heat from gas turbine exhaust and converts it into electric energy by means of a double closed loop system (see Figure 2). The thermodynamic cycle is based on an Organic Rankine Cycle (ORC) working with a hydrocarbon fluid (Cyclopentane C5). Heat from the turbine exhaust is transferred to a closed diathermic oil loop, which is used to heat an organic fluid loop. This lower temperature heat is then converted into useful work that can generate electricity. The system is similar to a conventional steam bottoming cycle except that the organic fluid drives a turboexpander that in turn drives the generator. The diathermic oil and the organic fluid allow low temperature heat sources 451ºC,7769 kg/s 1,75 bar to be exploited efficiently to Exhaust gas produce electricity over a wide 29ºC range of power output, from a Thermal oil few MW up to 17MW per unit. Exhaust (shell & tube) Cyclopentane circulation pumps Evaporator (brazed plate or shell and tube) Cond./evap. (brazed plate) Nitrogen blanketting HT expander (radial type) LT expander (screw type) R-245fa Expansion vessel hot oil storage 23ºC X = 1 X =.95 Recuperator Preheaters Vaporizer Superheater Hot oil pump The organic working fluid is vaporized and pressurized in the heat exchanger system by the application of heat transferred from the gas turbine exhaust stream. Then, the vapor expands in the turboexpander where the thermodynamic characteristics of the working fluid allow a dry expansion. FIGURE 2 Gas Engine heat recovery systems developed by GE Jenbacher /GE Global Research Center FIGURE 3 GT waste heat recovery systems developed by GE Oil & Gas / GE Global Research Center 3

5 Since the working fluid has not reached the two-phase state at the end of the expansion, its temperature at this point is higher than the condensing temperature. For this reason, the thermodynamic cycle efficiency is improved by means of a recuperator in which residual heat from the expander exhaust is recovered before it passes to the condenser. This higher temperature stream can be used to preheat the liquid cyclopentane before it enters the main heating units thereby reducing the amount of heat required from the hot oil and increasing the cycle efficiency. The fluid at the recuperator outlet is condensed using air-cooled heat exchangers at ambient conditions. The condensate is pumped back to the recuperator and to the heat exchanger system, closing the thermodynamic cycle. The heating and cooling sources are not in direct contact with either the working fluid, or the expander. For higher temperature applications, a high temperature thermal oil is used as the heat carrier and a regenerator is added to further improve the cycle performance. The selection of the working fluid is key in a Rankine Cycle. The main features of the ORegen working fluid are: Low freezing point and high temperature stability High heat of vaporization High density High molecular weight High critical temperature and pressure Low environmental impact No additional EHS considerations Readily available at low cost ORegen configurations In order to properly address the capability of the system to produce additional power over a wide range of output power and across a wide range of waste heat energy, the following system configurations have been analyzed along with the scaleability of the systems: Single exhaust heat source Multiple exhaust heat sources FIGURE 4 Single exhaust heat source In the single exhaust heat source configuration, the waste heat energy generated by the GT is delivered to the ORC system through a single Waste Heat Recovery Unit (WHRU) installed in the oil loop subsystem FIGURE 5 Multiple exhaust heat sources In the multiple exhaust heat source configuration the waste heat energy generated by several GTs is delivered to a single ORC system through multiple WHRUs (one for each GT) installed in the same oil loop subsystem (see Figure 5). The maximum capability of an ORegen unit is 17 MW; Figure 6 shows a typical ORegen installation. FIGURE 6 Typical ORegen system WHRU ORC Condenser WHRU #1 WHRU n Waste Heat Recovery Unit ORC Organic fluid system Condenser Working fluid selection The selection of the working fluid used for GT waste heat recovery is a key factor in the design of the system. GE s Global Research Center (GRC) in Munich, Germany made a selection of the working fluid during the initial development stage of the system. The study took into account the following key parameters: Ideal vapor pressure: Not too high: condenser cost/fan power grows with vapor pressure 4

6 Not too low: fluid should condense above atmospheric pressure to prevent air injestion High critical pressure allows greater expansion after boiling High molecular weight Flow rate inversely related to molecular weight of fluid Turbine cost grows with fluid volumetric flow rate Large enthalpy yield from expansion Large Δh correlates with vertical saturated vapor line Ideally, an organic material with: Condensation temperature close to t ambient at ambient pressure High critical pressure After eliminating many possibilities in a prescreening phase, the final assessment consisted of analyzing the following three potential fluids (see Table 1): R-245fa Thiophene Cyclopentane Fluid Property Boiling point (affects condenser cost) Freezing point (operability in cold environment) Molecular weight (increase lowers turbine cost) Velocity of vapor line (increase lowers condenser cost) Table 1 R-245fa Thiophene Cyclopentane ºC -38ºC -94ºC Fair Good Fair The analysis highlighted cyclopentane as the most suitable compromise (i.e., trade off between the boiling and freezing point, molecular weight, verticality of vapor line on T-S diagram) fluid for this type of application, giving acceptable to high performance in all categories. Cyclopentane s main thermophysical properties are: Boiling point: 121 F (49.3 C) Freezing point: -137 F (-94 C) Molecular weight: 7.1 To complete the working fluid assessment, an additional analysis evaluating the long term behavior of the fluid (see Figure 7) and fluid/metal contamination (see Table 2) were carried out. XDecomposition product Timeframe for which direct measurements exist Figure 7 worst-case scenario Concentration measurement 95% confidence interval The analysis of the long term behavior of the working fluid showed a residual risk due to cyclopentane decomposition. The decomposition of cyclopentane would lead to a negligible long term cycle performance loss of 2% over 1 years. Fluid metal corrosion evaluation results show no indication of cyclopentane-induced corrosion in either stainless steel or generic carbon steel. Penetration rate <=.5 mils/year Extrapolated curve fit Possible range of curves (given measured variability) ORC Performance after fluid decomposition Heat source: PGT25+ under ISO conditions Worst-case decomposition Expectation Best-case decomposition Intermediate fluid selection The purpose of the intermediate fluid is to transfer heat from the WHRU installed in the GT exhaust duct to the heat exchanger system, which heats the working fluid. The intermediate fluid (hot oil) is also employed to increase the safety of the system by avoiding the direct transfer of heat from the exhaust gas to the cyclopentane. The selection criteria for the hot oil take into account the following key factors: ORC net output Frosh (95% cyclopentane) charge = Years of operation with single fluid charge t 5

7 Corrosive Ferrous alloys Cast iron Gray Nickel Silicon Mild steel Austenitic stainless steels 32,34 321, ,317 ACI CN-2 2Cr-3Ni Martensitic stainless Copper Copper base alloys Brass 7-8Cu +Zn, Sn/Pb Brass 59-93Cu +Al, Zn/As Cupro- Nickel 66-88:11-33 Cyclopentane or Cyclopentadiene C F Percent concentration in water Average penetration per year Code Mils 1 inch/1 Microns < < { Some conversion factors Steel: mpy = lb/ft 2 /yr x 24.5 ipy x 696 x density = mdd g/m 3 /d x.144 density = ipy 1 micron =.3937 mil Parts per million =.1 g/liter Normal = 1 g equiv. liter (wl) X > Average penetration rate/yr compared to weight loss CDS mg/dm 2 /day g/m 2 /yr lb/ft 2 /yr CDS mg/dm 2 /day g/m 2 /yr lb/ft 2 /yr Aluminum Lead <3.79 <138 <.284 <15.75 <576 <.1178 <31.5 <576 <1.178 <37.9 <138 < , X >945.5 >345 >.71 X > >14,4 >2.945 Copper, Nickel or Iron Tantalum <11.9 <435 <.896 <23.6 <843 <.172 <23.6 <843 <1.72 <119. <435 < , X >297.5 >1875 >2.24 X >576.5 >21,75 >4.3 Table 2 Minimum pumpability temperature (TMIN) Operability range The operability range is defined as the interval between the Minimum pumpability temperature(t MIN ) and the Maximum operating bulk temperature (T MAX ). The wider the operability range, the more suitable is the fluid. The intermediate fluid selection was made considering three different potential fluids: Fluid A: T MIN = C t MAX = 345 C Fluid B: T MIN = -4 C t MAX = 315 C Fluid C: T MIN = -35 C t MAX = 33 C Fluid B was selected as the most suitable for this application since it does not require an oil heater or piping tracing (T MIN =-35 C), and is able to withstand sufficiently high temperatures so as to not cause cycle performance losses. The main characteristics of the selected intermediate fluid are: Minimum pumpability temperature: -35 C (-3 F) Maximun stability temperature: 33 C (625 F) Maximum (recommended) film temperature: 36 C (675 F) Autoignition temperature: 412 C (412 F) Temperature operability range: -35 C (-3 F)/ 33 C (625 F) Flash point: 12 C (248 F) Case study GE Oil & Gas and the GE Global Research Center developed the ORC application (see schematic in Figure 8) to recover electrical Pump power: Oil heater Pump power: Preheater Dtrm oil Boiler FIGURE 8 Case study schematic Superheater TE Aux power: Fan power: Recuperator Condenser energy from GT waste heat using, as a case study, a PGT25+ mechanical drive gas turbine installed in a pipeline compression station. This study shows the capability of the system to produce electricity using the energy that would normally be released into the atmosphere from the gas turbine exhaust gases. The system design philosophy was developed using cost and simplicity as driving requirements, and therefore, component redundancy was minimized. However, additional requirements can be taken into account (i.e., system reliability) by adding component redundancy with a consequent impact on CapEx. Air 6

8 The system configuration is the single exhaust heat source configuration implemented with the dual closed loop systems. The analysis was performed considering the design of the following ORC system main components: Pumps Waste Heat Recovery Unit Working/Intermediate fluid heat exchanger system Driver Machine Air Condenser Pumps Vertical pumps were selected for both the oil and cyclopentane loops: Hot Oil Pump: Vertical in-line single stage overhung (API61 OH3) (see Figure 9) Cyclopentane pump: Vertically suspended canned multistage pump (API61 VS6) (see Figure 1) Efficiency (%) Head (m) 97.5 NPSHR Efficiency Head Power Flow (m 3 /h) FIGURE 9 Hot oil pump NPSHR (m) Power (kw) FIGURE 1 C5 pump Waste Heat Recovery Unit The WHRU was designed based on the PGT25+ exhaust Stack data with a Tamb of 5 C. The WHRU main characteristics are: Bypass Heat recovered = approx. 3 MW Process fluid type: Diathermic oil Tube bundle composed of finned seamless tubes WHRU Casing: Carbon Steel internally insulated Efficiency (%) Head (m) 13 NPSHR Efficiency 6 5 Head Power Flow (m 3 /h) NPSHR (m) Power (kw) in Figure 11. The gas turbine exhaust enters the WHRU through a louver arrangement (multi-blades or single flap). The exhaust gas can be directed to the WHRU tube bundle, to the bypass stack or a combination of both. When the gases are diverted into the WHRU, the ORC system is warming up or running, while when the gases are diverted to the by-pass, the ORC is shutting down or cooling. Approx. Dimensions = 14 (L) x 4.9 (W) x 9 (H) meters Mechanical Design Code: ASME Sect. VIII, Div. 1 Since the WHRU is a cost-critical component for the ORC system, a sensitivity analysis was performed. The goal of the analysis was to evaluate the effect of the change of the DT approach on the system performance and component dimensions and costs. The analysis showed no significant advantages in reducing the DT approach; hence, a standard DT approach of 2 C was chosen (see table and graphic in Figure 12). Oil mass flow [kg/s] Oil heater minimum approach DT [C] Area increase estimate [%] Power gain % -2.7% % -1.% %.% %.7% DT approach 2ºC gas from GT hot oil Exchanged heat 49ºC 3ºC FIGURE 11 WHRU arrangement The WHRU (Waste Heat Recovery Unit) arrangement is shown FIGURE 12 WHRU sensitivity analysis 7

9 Working/Intermediate fluid heat exchanger system The heat exchanger system is composed of several hot oil/working fluid heat exchangers (preheaters, evaporator and superheater), and an additional heat exchanger (recuperator) installed to increase the cycle efficiency (see Figure 13 for preliminary sizing). FIGURE 15 Air condenser The main features of the condenser (see Figure 15) are: FIGURE 13 Heat exchangers Driver machine selection The selection of the generator drive machine for this ORC application considered two possibilities from the GE Oil & Gas product line (see Figure 14) Axial Expander Radial Expander FIGURE 14 ORC Driver machine (left: axial expander, Cooler Type: Full Welded Plug Type Materials: Carbon Steel Tube Length: approx 16 m Fin Type: Aluminum Embedded Fins Fin Number: approx 5 fins/m Fan Diameter: 4 m Case study results The analysis demonstrates the system capability for the production of electric power based on the assumptions listed below: GT Type: PGT25+ GT Application: MD/Pipeline Compression Station Design Ambient Temperature: 5 C (41 F) GT Design Load Condition:1% GT Load range: 5%-1% with the following results: center & right: radial expander) Both solutions were considered to be technically feasible and similar in terms of performance. The Radial Expander solution was considered to be preferable in terms of cost and manufacturing lead time. Air condenser To condense the working fluid, a GE Oil & Gas air condenser was designed specifically for this application. The condenser design conditions are: Tair 5 C Tin 68 C (slight superheat) Net Expected Power: 6.4 MW (Figure 16 shows a qualitative performance curve vs T amb ) Power min. air T FIGURE 16 Power vs. T amp Condenser size increase design T new new max. air T 5ºC design Tdesign T Case study business implications The value to the customer of ORegen includes both the sale of the electricity produced from the bottoming cycle, as well as the carbon credit. In addition, the customer may benefit 8

10 from the revenue of the additional gas produced by the process in case the power generated is used in an oil and gas plant. The case study payback analysis is based on the assumptions outlined in Table 3. Hardware ORegen & electric cabin $M 15 Installation $M 6 Cost of capital % 8 Tax rate % 35 Energy market price Energy market price AAGR $/MW % credit trading, which total to $2M. As cash out, there is the $15M cost of ORegen, the installation cost $6M, and the equipment maintenance cost $2.6M. The investment in this specific case study can be summarized by the following finacial indicators. Net present value (NPV) $5.3M Break-even point (BEP) 8 years Return on investment (ROI) 33% Internal rate of return (IRR) 12% Carbon credit $/ton 7 CO 2 production avoided ton/yr 3, $M 18 years 2 15 Customer value equation Energy production MWh/yr 47,654 ORegen power output Mwe 6.4 ORegen reliability/availability 85% Maintenance cost 6 year cycle $M 1.9 Expander overhaul.6 Heat exchanger overhaul.1 Gen set.2 Other 1 Electricity sales CO2 credit ORegen capex Installation cost Figure 17 Investor s net present value ORegen maintenance cost Investor NPV Cyclopentane cost $k/yr 18 Consumption ton/yr 12. Density kg/bbl 99.4 Barrel consumption bbl/yr 12.8 Unit cost $/bbl 15. Table 3 No government subsides or tax allowances were considered in the payback analysis, even though the governments of many countries have already enacted laws to stimulate such waste heat recovery projects to support the Kyoto Protocol or other environmental initiatives in the country. The payback of an industrial plant is usually considered as 15-2 years; in this case study, we assumed 18 years. Based on these assumptions, the revenue from the sale of electricity over the 18-year span, discounted at a rate of 8%, is $27M. In addition, there are the benefits of carbon 9

11 Conclusions The key features of ORegen vs. a typical water/steam cycle are summarized in the following table. Features ORegen Standard steam combined cycle Water free *** * Low maintenance *** * Power flexibility *** * Low investment *** * Plant simplicity *** * Reduced footprint *** * No replenishment of working fluid *** * Unmanned capability *** * Additional output ** * * Table 4 The flexible system architecture and reduced footprint allows ORegen to be easily retrofitted into existing plants. The system can handle the recovery of waste heat from the exhaust of small to medium size gas turbines. The following table represents the range of power output that could be recovered from the exhaust of the various GE Oil & Gas gas turbine models For more information Global Headquarters Via Felice Matteucci, Florence, Italy T F customer.service.center@ge.com Nuovo Pignone S.p.A. Americas Regional Headquarters 4424 West Sam Houston Parkway North Houston, Texas 7741 P.O. Box 2291 Houston, Texas T F ge.com/oilandgas The information contained herein is general in nature and is not intended for specific construction, installation or application purposes. GE reserves the right to make changes in specifications or add improvements at any time without notice or obligation. 21 General Electric Company All Rights Reserved Gas Turbine Model Gas Turbine Power (kw) Exhaust flow (kg/sec) Exhaust temp (ºC) Gas Turbine Efficiency (%) ORegen output (MWe) PGT25 23, PGT25+ 31, PGT25+ G4 33, MS51 26, MS52C 28, MS52D 32, Ms61B 43, Table 5 ORegen is a competitive alternative to the standard water/steam Rankine Cycle from a CAPEX and operability standpoint. In many cases, it is the only viable solution for the recovery of heat from the exhaust of a gas turbine in a mechanical drive application in which the load may dramatically change over time and where the plant may be sited where water is not available. 1