Validation of Ignition Reliability in SGT-750 Combustor for Diluted Natural Gas at Extremely Low Ambient Temperature

Transcription

1 POWER-GEN Europe, Cologne, Germany, June 27-29, 2017 Validation of Ignition Reliability in SGT-750 Combustor for Diluted Natural Gas at Extremely Low Ambient Temperature Jacek Janczewski - Advisory Combustion Expert Mats Andersson - Senior Combustion Expert Olle Lindman - Senior Expert Combustion Systems Magnus Persson - Combustion Expert Siemens AG Copyright Siemens AG, All rights reserved.

2 Abstract In order to address a broader range of gaseous fuels a combustion test was performed in the SGT-750 prototype unit on natural gas diluted with inert gases. At up to 50 vol% nitrogen and 40 vol% carbon dioxide content in natural gas the flame ignition and acceleration of the unit were demonstrated with satisfactory reliability. The engine tests were followed by complementary combustion tests of the SGT-750 original single burner combustor, carried out in a high pressure as well as in an atmospheric combustion test rig at conditions corresponding to extremely cold ambient (down to -60ºC). In the former case (high pressure rig) the combustion of CO 2 / N 2 diluted natural gas was studied for heat rates, air pressures and temperatures scaled to the engine operation at -60 C ambient across the SGT-750 operating line from idle to 100% load. It was shown that the combustor could be operated with preserved reliability and performance. This paper focuses however on a description of the latter case (atmospheric combustion rig) when the same test hardware was used with the process air cooled down to -60 C. Air flow rate to the rig was the same as at start-up conditions in the engine. A special air supply unit incorporating liquid air storage, steam driven liquid air evaporator and the flow and temperature control station was assembled for the test. The test was conducted for a variety of air flow rates in the temperature range of C and the start settings of fuel flow rates to the igniter and main burner in order to find the optimal ignition conditions. For the test purposes even the energy and the frequency of the spark exciter was altered. The test was run with natural gas and its blends of inert gases i.e. nitrogen and carbon dioxide. The fuel could be ignited at the power corresponding to SGT-750 start up conditions for the air temperature of -60 C with the igniter and the burner fed with pure natural gas and blends with up to 55 vol% nitrogen and 30 vol% carbon dioxide respectively. The engine and the rig tests have proven that SGT-750 can be started and operated within a broad range of ambient conditions (arctic climate) and low heating value fuel blends. This paper provides a description of the test setup and evaluation of the operational data for a single burner SGT-750 combustion system in the atmospheric combustion test rig for the gas blends at cold ambient conditions. Copyright Siemens AG All rights reserved. 2

3 Table of contents Abstract Introduction Test Setup Testing Results Conclusion...15 References...15 Permission for use...16 Disclaimer...16 Nomenclature ACR Atmospheric Combustion Rig AFR Air Fuel Ratio AFRst Stoichiometric Air Fuel Ratio CO 2 Carbon Dioxide DLE Dry Low Emission ER Equivalence Ratio FD Flame Detector HPCR High Pressure Combustion Rig ISO International Organization for Standardization LHV Low Heating Value LNG Liquefied Natural Gas N 2 Nitrogen NG Natural Gas O 2 Oxygen RPL Rich Pilot Lean TC Thermo-Couple WI Wobbe Index Copyright Siemens AG All rights reserved. 3

4 1. Introduction The SGT-750, shown in figure 1, is a high performance twin shaft gas turbine rated at 41 MW with 41.6% simple cycle efficiency (ISO) with a pressure ratio of 24 and a compressor discharge temperature of 490ºC. Figure 1: The SGT-750 twin shaft 41 MW engine The engine comprises a combustion system with eight cannular type combustors. The single combustor incorporates a Dry Low Emission (DLE) burner, a serially cooled can and a parallel cooled transition duct. The air flow scheme within the combustion system is shown in figure 2. Majority of the air leaving the compressor diffuser enters the cooling channel of the can and flows into the burner. In the burner the combustion air is distributed between two main passages of the swirl generator, the central pilot and the RPL (Rich Pilot Lean) igniter. Combustion stability in the partially premixed DLE flame is maintained by a central recirculation zone created by means of a vortex breakdown in the aft part of the can. The remaining part of the compressor air flows into the hot gas path downstream the flame through a plurality of small orifices in the double-skin wall of the transition duct. This design provides an effective cooling of the transition duct and allows achieving a desired temperature in the primary zone. Copyright Siemens AG All rights reserved. 4

5 Burner Can Transition duct Figure 2: SGT-750 combustion system There are numerous projects both on the mechanical drive and the power generation markets, where the SGT-750 power class gas turbines are in demand of wide range of fuel flexibility both in terms of limited WI (Wobbe Index) and fuel reactivity (process gas blends with high content of hydrogen/carbon-monoxide, high content of heavy hydrocarbons). Usually the primarily examined option of widening gaseous fuel specification for a new type of gas turbine is gases originating from Liquefied Natural Gas (LNG) processes. Such gases are often diluted with a high amount of inert components such as nitrogen and/or carbon dioxide. Some of the projects are located in areas with severe arctic climate which leads to a demand of proven operation under such conditions. The development strategy of how to simulate operation of gas turbine with both diluted natural gas and at very low ambient temperatures was established for SGT-750 in several steps: Simulation of ignition and startup on natural gas (NG)/N 2 /CO 2 blends and the process air temperature decreased to what is expected at true arctic conditions (-60 C). Atmospheric test envelope would be feasible to complete the test. Verification of the combustion system performance and operation at pressurized conditions where the air pressure and temperature would be adjusted to the operating line corresponding to the true arctic temperature constraints. The fuel composition is altered by mixing required amount of N 2 and CO 2 in NG. Final verification of content of the inert gases in NG on operation ability during start and loading in the SGT-750 engine. The two former cases were accomplished in the test rigs. Test with the process air temperature reduced down to -60 C was only feasible in the Atmospheric Combustion Test Rig (ACR) at Siemens test facility in Finspång, Sweden. The second test campaign was performed in the High Pressure Combustion Test Rig (HPCR) in Cologne, Germany. In both cases a full scale, original engine hardware of the single burner SGT-750 combustor was used. In the latter case, a full scale test of the SGT-750 engine was carried out in the Siemens gas turbine test facility in Finspång, Sweden. The test conditions were close to ISO which is typical for the Swedish summer weather. This paper gives emphasis to the test setup, execution and results obtained from the test campaign in the ACR. The results from the HPCR and the engine tests are briefly presented in the conclusions. The details of the tests can be found in [1]. Copyright Siemens AG All rights reserved. 5

6 2. Test Setup Siemens Industrial Turbomachinery in Finspång, Sweden has had a test facility for combustion purposes at close to atmospheric pressure since the 1990's. The facility is quite suitable for testing of components such as burners, combustor liners and auxiliary system used for combustion monitoring. The test rig offers air preheating, gas fuel mixing, optical access to the combustion confinement etc. The flow capacity of air, fuel and handling of hot exhaust gases was big enough to perform the ignition and startup test with a single burner, full scale SGT-750 combustor. The architecture of the SGT-750 combustion system incorporates eight separate combustors equipped with individual spark igniters and flame detection devices. The combustors can communicate with each other only through the air plenum in the central casing and the small annular cavity in front of the nozzle guide vane. Ignition and startup of gas turbine occurs usually at the pressure only slightly exceeding ambient pressure. Consequently, for the reasons above it was believed that the atmospheric test facility fulfills criteria for a cost-effective and reliable testing of ignition and flame detection. An image of the single combustor assembled in the ACR is shown in figure 3. Can Test vessel Transition Duct Emission Probe Burner Main FD View Port RPL FD/TC RPL Air inlet Exhaust Channel Figure 3: The SGT-750 combustor assembled in the ACR In order to fulfill the test conditions, the rig was supplied with gas fuel from a fuel mixing station where the fuel mixture with the required content of CO 2 / N 2 in natural gas was provided to three fuel lines i.e. Rich Pilot Lean (RPL) igniter, pilot and main nozzles, normally used during the ignition and startup phase of SGT-750. The flow rates of the NG, N 2 and CO 2 where measured with Coriolis flow meters to establish the required mass content of the inert gas, which was further recalculated to the volumetric concentrations and the actual heating value of the fuel. The ignition heat power at the moment of ignition was controlled by three governing valves, one for each of the fuel lines by means of the total fuel flow rates measured with the Coriolis meters. Copyright Siemens AG All rights reserved. 6

7 In order to provide the combustion air to the test rig with the required O 2 / N 2 content and temperature a special cold air supply unit was assembled. The air supply unit consisted of Liquid air storage with capacity of 20,000 kg of liquid mixture of oxygen and nitrogen with initial concentration of O 2 / N 2 at 23 / 77 weight percent at 5 bar pressure and -220ºC i.e. 20 degrees below the boiling point Liquid air evaporator with a steam driven heat exchanger to control air temperature in gas phase Diesel fired steam generator Air flow governing valves and orifice plate for control of the air flow rate in the gas phase The air supply unit was designed and provided by the Swedish company AGA possessing the sound knowledge and unique experience in the area of gas handling techniques. Since there was a risk of separation of the gas components in the liquid air storage, the oxygen content was continuously monitored with an oxygen analyzer through a sampling probe installed in the exhaust. The ignition power in SGT-750 provides a very low energy to the exhaust gas even when burning, so as an extra measure against freezing of the exhaust channel and the stack, some fresh air at ambient temperature was injected downstream the test vessel. A schematic of the entire test facility for test with the cold air and gas blends is shown in figure 4. Figure 4: Schematic of the ACR Copyright Siemens AG All rights reserved. 7

8 The liquid air storage shown in figure 5, had a capacity which was enough for ~30 hours rig operation at SGT-750 start conditions. The major challenge for tests with liquid air is the risk of separation of the gas components in the liquid air storage. Usually for combustion tests with air at extremely low temperature in laboratory scale oxygen and nitrogen are supplied through separate lines to a mixing unit just before the combustor. This makes the air supply facility quite complex, as it was shown in [2]. For this reasons Siemens has chosen another option with direct evaporation of artificially prepared liquid air. The vendor of the entire air supply unit was the Swedish company AGA. The requirements for the test were as follows: Air flow rate: controlled within the range of g/s, accuracy 20 g/s Air supply temperature to the rig: controlled within the range of ºC, accuracy 5 degrees Oxygen content in the process air shall be kept within the range of vol%. These test conditions were fulfilled successfully. Air storage Evaporator Steam generator Figure 5: View of the air supply unit The fuel mixing unit comprised liquid nitrogen and compressed carbon dioxides storage. The inert gases were mixed in the mixing station with natural gas before the gas blend with desired WI was fed into the burner. The set point for the actual test conditions was however the volumetric concentration of the inert gas, which was established according to measured mass flow of the gas blend components. Some details of the gas mixing station can be seen in figure 6. Copyright Siemens AG All rights reserved. 8

9 NG line Governing valves N 2 line CO 2 line Flow meters Mixing device Figure 6: View of the fuel gas mixing station 3. Testing The objective for testing was to verify ignition ability of the SGT-750 single burner combustor at the simulated arctic conditions. The test comprised several stages initiated with a reference test with standard fuel (NG) at ambient conditions close to ISO. Furthermore, ignition of the main flame in the test rig was monitored with the original engine equipment and it was supposed to fulfil the same criteria of flame detection as it happens in the real engine i.e. follow up the following sequence: Spark igniter is on Fuel to the RPL igniter is fed If flame in the RPL igniter is indicated by a built-in TC (condition for the temperature increase are fulfilled) Fuel to the main injector is being fed The main flame light is indicated by the main flame detector (flame eye built-in in the burner) If so, proceed to increase of the fuel to the burner according to the fired power and fuel split schedule If not, interrupt fuel feeding to the burner for all the fuel lines If the RPL temperature increase does not fulfil the criteria suspend the start sequence and try again For the arctic conditions (temperature range of the process air in the range of ºC) and the diluted natural gas another test objective was to find optimal start settings for reliable ignition with respect to: Air flow of the process air in the engine controlled by the ignition rotor speed of the compressor Equivalence ratio (ER) of the fuel fed to the RPL and the main burner so called ignition power Vary fuel split pilot/main at need Copyright Siemens AG All rights reserved. 9

10 In terms of control parameter of the process it was assumed that the air temperature will be first reduced down to -60ºC and the air and fuel flow to the burner will correspond to the present engine settings of the ignition power and the compressor rotor speed. This would set the equivalence ratio in the RPL igniter and the main burner to the same level as it is presently used to start the SGT-750 for ambient temperatures in the range of ºC. In case of difficulties to light the RPL's and the main flames both the air flow and the fuel ignition power would be varied. The second option which could be used in case of failed ignition of the given fuel would be a stepwise increase of the process air temperature by 10 degrees at the time until the flame would be lit. The test matrix in table 1 shows how the grade of natural gas dilution with nitrogen and carbon dioxide was altered during the test. NG Air temperature [ C] 15/-30/-60 Table 1: Matrix of fuels used for the ignition test CO2 blend [CO2/(CO2 + NG)] N2 blend [N2/(N2 + NG)] 15/-60 15/-60 Mole content [vol%] Weight content [mass%] LHV [MJ/kg] WI [MJ/m^3] In the beginning of the testing it was found that the thermal insulation of the air hoses between the air evaporator and the test vessel as well as the vessel itself was not sufficient enough to reach -60ºC due to natural heating of the cold air from the ambient. A new, more effective thermal insulation was then applied, as it is shown in figure 7. As it can be seen, there are heavy ice deposits on the noninsulated surface of the vessel and the front face of the burner. Figure 7: The test rig in operation during the ignition test. The burner and the air hose are shown to the right. The rig and the test hardware were instrumented with mass flow meters, thermocouples and pressure transducers in order to control the process data. Standard SGT-750 combustor monitoring equipment was installed in the test hardware. The process and the rig data were followed on line and recorded in the data acquisition system with sampling resolution of one second for further post processing and evaluation of the test results. Two examples of the time trends illustrate successful ignition for the most excessive test cases when the air inlet temperature to the test vessel reached -60ºC. In the former case in figure 8, NG was diluted with 30 vol% of CO 2. In the latter case shown in figure 9 a NG blend with 60 vol% N 2 was tested. In the both cases the ignition sequence was tested twice in order to secure repeatability of the successful ignition. Copyright Siemens AG All rights reserved. 10

11 Main FD 0-10V Air flow g/s CO 2 flow 0-20g/s NG flow 0-20g/s O 2 in air 0-25vol% RPL temp C Air temp C Figure 8: Ignition test with 30 vol% carbon dioxide The brown curve indicates temperature of the RPL igniter the increasing trend shows that the RPL flame was lit. When the RPL ignition criterion is accomplished (temperature level and gradient) the main fuel with the inert gas is injected and the main flame is lit, followed by sudden rise of the FD signal (the red curve). The figures in boxes show names of the recorded channels as their ranges in the plot area. O 2 in air 0-25vol% Main FD 0-10V N 2 flow 0-20g/s Air flow g/s Air temp C RPL temp C NG flow 0-20g/s Figure 9: Ignition test with 60 vol% nitrogen. It is worth to notice that the oxygen content in the process air was in the range of vol%, which was acceptable for the test purposes. Copyright Siemens AG All rights reserved. 11

12 4. Results The test results are summarized in form of charts where a positively confirmed ignition of either the RPL or the main flame are represented by unity (1), while a failed ignition is depicted as zero (0) on the vertical axis of the diagrams. The minus one (-1) represent the cases which were omitted during the testing for different reasons. The ignition outcome is plotted versus ER of the air/fuel mixture for the RPL and for the main burner and the molar content of the inert gas in the fuel blend. The ER is defined as where AFRst is the air-fuel ratio at stoichiometric conditions M air is the air mass flow in the air/fuel mixture entering flame M fuel is the fuel mass flow entering flame In this case the ER is defined for the local air inflow controlled by internal air split in the combustor for the main flame and the internal air split in the burner for the RPL. The diagrams treat separately the reference cases for ambient conditions close to ISO (figures 10 & 12) and the test cases with the air temperature reduced to -60ºC (figures 11 & 13). Copyright Siemens AG All rights reserved. 12

13 RPL ignition window for CO 2 gas blends - ISO RPL ignition window for N 2 gas blends - ISO 0-failed; 1- flame on Equivalence Ratio [-] vol% 30vol% 20vol% NG NG vol% vol% vol% failed; 1- flame on Equivalence Ratio [-] vol% 40vol% 30vol% 20vol% NG NG vol% vol% vol% vol% Figure 10: Ignition ability of the RPL for NG and its blends with CO 2 and N 2 at ISO conditions RPL ignition window for CO 2 gas blends -60 C RPL ignition window for N 2 gas blends -60 C 0-failed; 1- flame on Equivalence Ratio [-] vol% 40vol% 30vol% 20vol% NG NG vol% vol% vol% vol% failed; 1- flame on Equivalence Ratio [-] vol% 40vol% 30vol% 20vol% NG NG vol% vol% vol% vol% Figure 11: Ignition ability of the RPL for NG and its blends with CO 2 and N 2 at T air ~ -60 C It can be seen that the operating window in regard to ER where the RPL flame can be ignited is relatively wide for pure natural gas and shrinks continuously towards increasing share of the inert gases in the gas fuel blend. However even for the highest tested contents of CO 2 and N 2 (48 and 55 vol% respectively) there is still a possibility of reliable ignition. It can be also seen that the range of the ER's where the successful ignition is still possible can be made even wider for the dilution grade up to 40 vol%, than it was for the pure NG. The reason of this is possibility to control the inlet air flow which was varied between 400 and 650 g/s. For a given ER it does mean that the ignition power (fuel flow) was altered to find optimal ignition conditions. This was not the case for the reference run at ISO conditions, where only the standard engine settings were applied. Copyright Siemens AG All rights reserved. 13

14 Main flame ignition window for CO 2 gas blends - ISO Main flame ignition window for N 2 gas blends - ISO -1 - N/A; 0-failed; 1- flame on Equivalence Ratio [-] 48vol% 40vol% 30vol% 20vol% NG NG vol% vol% vol% vol% N/A; 0-failed; 1- flame on Equivalence Ratio [-] 55vol% 40vol% 30vol% 20vol% NG NG vol% vol% vol% vol% Figure 12: Ignition ability of the main flame for NG and its blends with CO 2 and N 2 at ISO Main flame ignition window for CO 2 gas blends -60 C Main flame ignition window for N 2 gas blends -60 C -1 - N/A; 0-failed; 1- flame on vol% 40vol% NG 20vol% 30vol% -1 - N/A; 0-failed; 1- flame on vol% 40vol% NG 20vol% 30vol% Equivalence Ratio [-] Equivalence Ratio [-] NG vol% vol% vol% vol% NG vol% vol% vol% vol% Figure 13: Ignition ability of the main flame for NG and its blends with CO 2 and N 2 at T air ~ -60 C It was proven that the successful ignition of the main flame subsequently to ignition of the RPL is feasible for the dilution grade of up to 40 vol% for CO 2 blends and up to 55 vol% for N 2 blends for the reference case at ISO conditions (figure 12). When the process air temperature was reduced to the simulated arctic conditions the ignition window reduces compared with the reference case (figure 13). It can be seen that the ignition with the CO 2 blend is achievable for wider range of the startup conditions with respect the ER than for the N 2 blends, however again the ignition of the CO 2 blend failed when the inert gas content exceeded 30 vol% while a successful ignition was still possible for 55 % of the N 2 blend. For the N 2 diluted fuel the safe ignition windows moves towards lower equivalence ratios. Copyright Siemens AG All rights reserved. 14

15 5. Conclusions The fuel flexibility test campaigns extensively performed in 2016 have proven the SGT-750 and its combustion system to be very tolerant to variation of fuel quality at various ambient conditions. The test campaign in the Atmospheric Combustion Rig, reported in this paper, was carried out with the original setup of the SGT-750 single burner combustor and the flame monitoring devices. In order to simulate the artic start up conditions in the engine, the test facility was equipped with the air supply unit where the process air temperature could be reduced to -60ºC, which satisfied the required test conditions. It was verified that the ignition capability and reliability at artic conditions is satisfactory for natural gas blends containing up to 55 vol% of N 2 and 30 vol% of CO 2. The startup settings tested in the rig for a variety of fuel compositions can be transferred and applied in the engine. Further verification of the operation envelope at arctic conditions, not reported in this paper, was verified for the SGT-750 single can combustor in a high pressure test rig where the air pressure and preheat temperature were adjusted to simulate engine acceleration up to base load at arctic conditions [1]. Finally, engine operation of the SGT-750, including ignition, start and transient load changes was successfully proven using gaseous fuels containing various amounts of nitrogen and carbon dioxide. The engine was operated on natural gas blend with over 50 vol% of N 2 and 40 vol% of CO 2. Ignition and start capability were verified with satisfactory result on natural gas diluted with up to 34 vol% of N 2 and 20vol% of CO 2. These results are reported in [1]. All the tests both in the combustion rigs and in the engine were performed with the standard combustor hardware valid for a commercial SGT-750. The results achieved for the inert gas fuels have proven wide acceptance limits of fuel flexibility in the SGT-750. References [1] Lindman, O., Andersson M., at al.; GT , SGT-750 fuel flexibility: engine and rig tests, ASME Turbo Expo 2017, Charlotte, June [2] Pucher, G., Allan, W.D.; GT , Turbine fuel ignition and combustion facility for extremely low temperature conditions, Turbo Expo 2004, Vienna, June Copyright Siemens AG All rights reserved. 15

16 Permission for use The content of this paper is copyrighted by Siemens and is licensed to PennWell for publication and distribution only. Any inquiries regarding permission to use the content of this paper, in whole or in part, for any purpose must be addressed to Siemens directly. Disclaimer This document contains statements related to our future business and financial performance and future events or developments involving Siemens that may constitute forward-looking statements. These statements may be identified by words such as expect, look forward to, anticipate intend, plan, believe, seek, estimate, will, project or words of similar meaning. We may also make forward-looking statements in other reports, in presentations, in material delivered to shareholders and in press releases. In addition, our representatives may from time to time make oral forwardlooking statements. Such statements are based on the current expectations and certain assumptions of Siemens management, of which many are beyond Siemens control. These are subject to a number of risks, uncertainties and factors, including, but not limited to those described in disclosures, in particular in the chapter Risks in Siemens Annual Report. Should one or more of these risks or uncertainties materialize, or should underlying expectations not occur or assumptions prove incorrect, actual results, performance or achievements of Siemens may (negatively or positively) vary materially from those described explicitly or implicitly in the relevant forward-looking statement. Siemens neither intends, nor assumes any obligation, to update or revise these forward-looking statements in light of developments which differ from those anticipated. Trademarks mentioned in this document are the property of Siemens AG, its affiliates or their respective owners. TRENT and RB211 are registered trade marks of and used under license from Rolls-Royce plc. Trent, RB211, 501 and Avon are trade marks of and used under license of Rolls-Royce plc. Copyright Siemens AG All rights reserved. 16