Experiences in Commissioning of a 1MWe Solar Thermal Power Plant in Gurgaon

Transcription

1 Experiences in Commissioning of a 1MWe Solar Thermal Power Plant in Gurgaon N.G.R. Kartheek 1*, Deepak Yadav 1, Rangan Banerjee 1, J.K. Nayak 1, Santanu Bandyopadhyay 1 Shireesh B. Kedare 1 1 Department of Energy Science & Engineering, IIT Bombay, Mumbai, India * Corresponding Author. Tel: , solarthermaliitb@gmail.com Abstract: Indian Institute of Technology Bombay (IITB) is implementing a 1 MWe grid connected solar thermal power plant project in the campus of Solar Energy Center, Gurgaon. IITB has done conceptualization, design, engineering, procurement, construction, commissioning, operation and maintenance of power plant. This plant is unique as it combines a direct and indirect steam generating solar field. Though site preparation started in January 2010, the construction phase began from January 2011 and the plant was commissioned in June It is noted that relatively limited experience in commissioning of solar thermal power plant is available in public domain. This paper presents practical experience in commissioning of the solar thermal power plant. Problems encountered with equipment operation, engineering errors committed, difficulties in system level commissioning especially operation of the plant in winter, site related operational issues and solutions met to solve these problems are discussed. The rich experience from commissioning will guide the installers for smooth commissioning of the solar thermal power plant. Keywords: Solar Thermal, Power Plant, Commissioning, Practical Experience, Equipment 1. Introduction The global demand for energy and more specifically clean energy is growing rapidly. With growing energy demand and green-house gas emission, concentrating solar power (CSP) is considered as one of the promising option and has invited wide attention [1]. India being a tropical nation is blessed with abundant solar resource and has potential for utilizing this solar energy for electricity generation. The government in India has launched an ambitious target of 20 GW of electrical power generation through solar installations, half of it through solar thermal route. It is noticed that there is very limited experience with solar thermal power generation in India, especially in technology development and system integration. With the above consideration, Ministry of New and Renewable energy (MNRE) has sanctioned a project to Indian Institute of Technology Bombay on Development of a MegaWatt-Scale National Solar Thermal Power Testing, Simulation and Research Facility with an intention to promote solar thermal power technology in India. The scope and objectives of the project are as follows. Establishment of a national research facility on solar thermal power (1MWe grid interactive) Establishment of test facility for component and system characterization Development of simulation facility for future scale-up of plant capacity [2] It is expected that the facility would facilitate research and development for cost reduction of CSP (Concentrated Solar Power). 2. Plant Description Solar thermal power technologies use concentrating mirrors to focus solar radiation on the receiver to directly generate steam or heat oil, which in turn generates steam in a heat exchanger. The generated steam is then used to drive a steam turbine in a conventional Rankine cycle. Out of four prevalent CSP technologies viz. Parabolic Trough, Fresnel Reflector, Central Tower and Dish, the power plant built uses a combination of the first two. Figure1 shows the schematic of process flow for the power plant. The parabolic trough field generates hot oil (therminol VP-1) at 390 C, which is fed into the heat exchanger. Simultaneously, the linear Fresnel reflector solar field generates steam at 40 bar (g), C that is added to vapor space of the steam generator. The saturated steam from the steam generator is superheated to 350 C in a superheater to run a steam turbine generating 1MWe peak at a design Direct Normal Irradiance (DNI) of 600 W/m 2.

2 Fig. 1: Schematic of process flow for the solar thermal power plant There were many vendors supplying individual component of the solar power plant, however there were no indigenous turnkey developers of solar thermal power plant in India. Indian Institute of Technology Bombay has taken initiative in this scenario for integration of components to develop a solar thermal power plant. During commissioning of the plant many issues of varying nature were faced. This paper is an attempt to document these problems so that project developers may benefit from it. On a broad basis the issues have been classified as operational, equipment and system problems based on the nature of experience gathered. Figure 2 shows a tree-diagram of the problems experienced during commissioning of the plant. Though site preparation started in January 2010, the construction phase began from January 2011and the hot commissioning activity began in October The problems encountered during commissioning of the plant are shown in a chronological order in Fig 3. Fig. 2 Classification of commissioning experience Fig 3: Commissioning problems in chronological order

3 3. Operational Problems During commissioning of the plant, a few problems occurred during operation of the plant. These problems were identified and solved at site. A detailed description of the problems is done in this section. 3.1 Leakages in the superheater The steam from the boiler is fed into a superheater and then to the turbine. On March 10 th, 2013, at a steam pressure of 35 kg/cm 2 (g), the super heater started to show leakage as shown in Figure 4. Steam Leakage Fig. 4: Superheater Leakage The exchanger operated fine till 35 kg/cm 2 (g) pressure, however once it reached above 35 kg/cm 2 (g) pressure the leak started and it became more intense with the increasing pressure. The exchanger was tested many times however the problem persisted. The leakages were observed because the bolts loosened due to high temperature. The bolts were tightened at appropriate torque using a torque spanner and the problem was solved. 3.2 Receiver tube leakage and receiver glass window breakage of the LFR system The operating pressure and temperature conditions of the LFR system is 42 kg/cm 2 (g) saturated. The receiver tubes are welded and the material of construction is Stainless Steel Schedule 40. The tubes were hydro-tested at 120 kg/cm 2 (g). However, when the tubes were subjected to 50 kg/cm 2 (g) saturated steam pressure condition during operation, leakages were observed in 12 welding joints as shown in Figure 5(a). The receiver tubes are placed at a height of 13 m from the ground and needed spot welding. Considerable expenditure was incurred on re-welding the tubes as a special crane was arranged and welding a single joint took 8 10 hours. In another incident, the focus of the LFR system was partially shifted from the tubes to the enclosure because of interruption in power supply to the tracking motors. As a result thermal stress was induced in the metal and the receiver glass window was broken as shown in Figure 5(b). Receiver Glass Window Breakage Receiver Tube leakage Fig. 5(a): Leakage in the receiver tube Fig. 5(b): Breakage of the receiver window glass

4 3.3 Water entry in instrumentation air line Demineralisation plant (DM) requires air supply for proper mixing of resin in mixed bed unit. During commissioning, water of the DM plant entered into the air line. As a result electro-pneumatic positioners of four control valves were damaged. The problem was solved after installation of an NRV in the compressed air line at the inlet of the DM plant. 3.4 Dry run of the boiler feed pump The Boiler Feed Pump (BFP) used in the power plant is canned motor type that should not run dry for more than 2 seconds. The BFP has suction from the deaerator and trips when the level in the deaerator falls below a minimum value L min. When the deaerator was being commissioned, it was observed that the BFP ran at full RPM but the water level in the boiler did not increase. Multiple runs were taken but the problem persisted. The plant was shut down and the suction line was checked. It was found that there was no water in the deaerator even though the level transmitter in the deaerator showed sufficient water. The level transmitter was checked and it was concluded that the transmitter malfunctioned due to air ingress. Fortunately, the pump did not get damaged and the level transmitter was recalibrated. To prevent any further problem in future, the BFP will be tripped with signals not only from the level transmitters but also from level switches in the deaerator. 4. Equipment Problems During plant commissioning, a few problems were encountered with a few equipment of the plant. The errors in specification of the equipment have shown up as problem during commissioning of the plant. A brief discussion about the problems encountered, experience and correction made to solve the issue is done in this section. 4.1 NRV at BFP exit Failure The Heat Exchanger was pressurized for the first time on October 1 st, 2012 using steam generated from the parabolic trough solar field. Subsequently, steam was being generated and blown to the atmosphere to clean the turbine inlet steam line. A schematic diagram of water flow from deaerator to heat exchanger is shown in Figure 6. Fig. 6: Process flow from deaerator to steam generator The BFP is operated based on the water level in the steam generator. When the Steam Generator (SG) water level is L max, the BFP is tripped and is switched on when the level reaches L min. On October 27, 2012 at about 02:00 pm when the heat exchanger was being pressurized (SG Pressure - 25 kg/cm 2 (g), Hot oil temperature C), smoke was observed from the running BFP. At this moment, the BFP was not running as there was sufficient water level in the SG. When the pump started smoking, vibration was also noticed in the deaerator. Following actions were taken immediately to prevent possible damage to equipment. The boiler feed pump was isolated by closing the outlet gate valve. Vent valves were opened to decrease the SG pressure. Solar field was defocused. Heat Exchanger was bypassed to limit the steam generation. It was noticed that the BFP suction line from deaerator to BFP was very hot and water temperature in the deaerator increased from ambient 30 C to 42 C and the pipe supports on the suction line were dislodged. The vibration sound in the deaerator was due to back flow of the steam from the SG and smoke from the pump was

5 steam. Since there was backflow of steam and hot water from SG to deaerator, it can be concluded that NRV 001 BFW at BFP 001 exit failed. The NRV 001 / 002 BFW are welded on a vertical line while upon checking it was observed that the NRVs are suitable for installation only on a horizontal line. The problem occurred because of incorrect specification of the NRV and installation. The NRVs have now been replaced with NRVs mountable in vertical lines. 4.2 Leakages in instrument stub connections on heat exchanger Temperature Transmitters have been fixed on oil lines going into and out from the heat exchanger. The vapor pressure of therminol VP-1 at 257 C is 1 bar (a) [3]. As temperature increases, the vapor pressure of the therminol VP-1 increases. The pressure of the therminol VP-1 in the circuit should be more than the vapor pressure and has to be maintained using nitrogen blanketing on the pressure vessels. During operation of plant when the HTF was heated to temperature above 257 C, leakage was observed from temperature transmitter stubs. Upon inspection it was found that HTF vapor was leaking from the screwed connection between the stub and the nipple of the temperature transmitters. The plant was immediately put into shut down mode to prevent further loss of HTF. As shown in Figure 7(a), the end connections of the temperatures transmitters stub were screwed. Welding on the screwed connections with HTF in the line was not considered as auto ignition temperature of the HTF is 621 C whereas the arc temperature of the welding is about 1000 C. Draining oil from the circuit was also not an option because the circuit was charged with approximately 20 Tonne of oil. Isolating the heat exchanger and locally removing the oil also meant removal of considerable volume of oil in the heat exchanger. Moreover, the valves isolating the exchanger passed a certain amount of oil due to which the exchanger can never be kept completely dry from oil. A vendor was identified to do online sealing of the stub. As shown in Figure 7(b), a metal clamp was put around the stub, the space between the clamp and the stub was injected with a sealant. However, vendor did not reveal the properties and details of the chemical used [4]. The leakage is now arrested and plant is working fine. Lesson learnt from the problem is that the HTF circuit should have minimum flanged and absolutely no screwed connections. Normally, complete welded connections are preferred for HTF loop. Leakage from Screwed connection Fig. 7(a): Temperature Transmitters showing Leakages Fig. 7(b): Temperature Transmitters with sealant 4.3 Leakages in compressed air piping Compressed air is required in power plant for operation of control and pneumatic valves as shown in Figure 8. The desired air pressure for operation of these valves is 5 7 kg/cm 2 (g); as a result the discharge pressure at the compressor is about 7.5 kg/cm 2 (g). The compressed air piping network at the project site is made of GI with threaded end connection. Ideally, there should not be any pressure drop from the compressor storage tank discharge to valve inlet. However, it was observed that gradually the pressure at the valve inlet started to drop to 1.5 kg/cm 2 (g) even though the pressure in the air storage tank was 7.5 kg/cm 2 (g). The reason is leakages from the threaded ends of GI pipes. There was significant amount of leakages and for continuous supply of air, the piping network needed to be tightened routinely. The lesson learnt is that piping network for compressed air line should be completely welded to minimize the leakages. The entire network of compressed air will be replaced by MS welded pipes when the plant will be shut down for longer duration in winter.

6 Fig. 8: Schematic sketch of compressed air flow 4.4 Communication problem between level I and level II As discussed earlier, the solar thermal power plant designed by IIT Bombay has two solar fields viz. Parabolic Trough and LFR. These two fields work independently from their respective level II PLCs. The hierarchy of level I and level II is as shown in Figure 9. Fig. 9: Control System Hierarchy for the solar thermal power plant The level II controls of the LFR and trough system were designed to receive command/overwrite from centralized level I system. This required interface of level I with level II of both solar fields. Level I system has port 502 / 503 for communication with level II of solar fields. The LFR system has a SCADA screen of its own while the trough system doesn t have an independent operating SCADA screen and can be operated from the Level I screen only. As a result, port 502 was allotted to trough field level II; it was however noticed that LFR level II can only communicate with port 502 and not 503. Hence, LFR field could not be linked with level I and is currently operating independently. Trough system can run with port 503 but will need some hardware modifications. The issue will be taken up during long term shut down of the plant. The lesson learnt is to check for compatibility issue during designing of the PLC system especially when working with multiple level systems. 4.5 Passing of steam from steam system valves Before putting steam into the turbine, the turbine inlet steam line has to be cleaned completely by blowing superheated steam at rated pressure. For increasing the pressure, the steam valve at the boiler exit has to be closed. During the blowing of steam, it was noticed that the main steam pneumatic MSV valve was passing some amount of steam even though the valve was closed and the following issues were observed while evaluating the problem: 1. The MSV is an ON / OFF valve. As MSV passing some amount of steam, the peak pressure could not be reached in the steam generator. These valves also did not have a manual handle for closing, as result the commissioning work was hindered. 2. There are no manual valves in the steam line that could cover for the above mentioned problem of pneumatic valves. 3. The MSV being an ON/OFF valve, desired degree of superheat was obtained by opening the steam traps and closing the MSV. This problem could have been avoided with the MSV being a control valve.

7 The lesson learnt is that the MSV should be a control valve with a manual handle. The main control valve should have isolation valves and a bypass globe valve to manually control the flow in case of problem with the control valve. 5. System Problems Power and water are two vital resources for construction and commissioning of a power plant. The plant has a borewell facility in its boundary and a dedicated power line to the substation to evacuate power from the plant. The same line will be used for meeting the electricity load of the plant in case the plant is not evacuating power. However, during the construction phase of the plant this line was not available and the power requirements were being met by a tapped connection from the nearby substation. However, due to heavy load-shedding this line proved to be unreliable. This resulted in problems that are discussed in this section. 5.1 HTF freezing The crystallization temperature of HTF Therminol VP1 is 12 C. The plant, located in Gurgaon experiences minimum temperature much below 12 C in winter. Figure 10 shows the ambient temperature profile at the project site from midnight to 08:30 am on February 24 th, It can be observed that for most part of the night temperature is below 12 C. The plant runs in antifreeze mode during winter where the oil is continuously circulated in the solar field and heated to 40 C by an inline heater and electrical heat tracing provided at the pump suction and discharge lines. Ambient Temerpature( o C) :00 AM 12:30 AM 1:00 AM 1:30 AM 2:00 AM 2:30 AM 3:00 AM 3:30 AM 4:00 AM 4:30 AM 5:00 AM 5:30 AM 6:00 AM 6:30 AM 7:00 AM 7:30 AM 8:00 AM 8:30 AM Time (hr: min) Fig. 10: Ambient temperature profile at the project site on February 24 th, 2013 The HTF pump was being used to circulate oil in the solar fields and pipe lines. However, due to unavailability of electrical power, a DG set was used to run the plant in antifreeze mode periodically depending on the reading of the temperature gauge at pump discharge. When the oil temperature in the gauge showed below 35 0 C, main HTF Pump and electrical heating system was switched on. This activity was performed routinely for more than a month. However on December 31 st, 2012 evening when the plant operator went to start the pump and heat tracing system, it was noticed that the pump started, went to 150 rpm and the speed dropped to 9-10 rpm. The same procedure was tried again next morning, however the problem persisted. It was observed that the pressure readings at the pump discharge and solar field outlet were 12 bar (g) and 3.2 bar (g) respectively. It was concluded that HTF was crystallized in the circuit and the pump could not circulate the oil in the circuit. The main HTF pump was attempted to run in February when the ambient temperature was more than 20 C in the day. However after multiple runs on February 15 th, 2013 the seal in the main HTF pump failed and oil started leaking from the seal pot. Similar problem was experienced when one storage tank pump was attempted to operate. The oil freezing occurred because the HTF Pump was run based on the observation of one temperature gauge. Since the Level II PLC of the trough system was not working at that time, temperatures in the field and at other locations could not be monitored regularly. As a result, oil circuit crystallization anywhere in the circuit implied that the pump could not be run. This problem could have been avoided with running the system for 24 hours but it was not done in order to save on diesel cost. The seals of the main HTF pump and storage pumps were replaced and the pumps were tested after the oil unfreezing in March. It is also important to monitor the oil temperature at all locations in the circuit. It is recommended that the system should run in anti-freeze mode round the clock during winters.

8 It may be noted that the oil loop consists of storage and main HTF pumps (2 + 2) (Figure 1). Following problems were noticed with respect to anti-freeze mode. 1. The pump seals do not have anti-freeze protection. As a result only the inline pump is protected from the oil crystallization while the oil in the seals of stand-by pump crystallizes. 2. Two storage tank pumps are used only when oil is to be pumped into the circuit. Hence they cannot be run in antifreeze mode because they do not form a loop unlike main HTF Pump. So with the current system and pump seal design, oil in the seals of three out of four HTF pumps crystallized. Suitable solution will be sought from the pump supplier for storage tank pumps and stand-by main HTF pump. 5.2 Uninterruptible power system (UPS) and solar tracker failure UPS system is installed for emergency backup to the PLCs and weather station. A standalone UPS system sustained the load for 6 8 hours. However, since the grid power was not available and DG system ran intermittently for anti-freeze mode, the solar tracker (measuring the radiation data) used to get off-set daily. As a result the worm gear in the tracker failed and the tracker stopped working. The UPS was being used for larger duration while it was meant to be operated only for a few minutes in-case of grid failure. This resulted in decreased back-up period of the UPS. The problem could be sorted out only after continuous power back-up. 6. Summary This paper describes the practical experiences and problems in installing a MWe-scale solar thermal power plant in India. The experiences have been classified into operational, equipment and system problems. Operational problems related to leakages in superheater, receiver tube leakages and receiver glass window breakage of the LFR system, water entry in instrumentation airline and dry run of BFP were observed. A few operational problems can result in equipment damage though it did not happen in this project. Problems like receiver tube and heat exchanger leakage were identified and rectified at the project site. Water entry in the instrumentation air line resulted in malfunctioning of 4 control valves. Equipment problems related to NRV failure, leakages in instrument stub connections and compressed air piping, communication failure between level I and Level II and passing of steam through MSV were observed. The NRV failure at BFP exit occurred because the horizontal line NRVs were fixed on a vertical pipe line. A few specific lessons learnt from the equipment problems are that temperature transmitter stub connections on the HTF loop and compressed air piping network should have welded connections and to check the compatibility of different PLCs during design phase. In turbine inlet steam line, the MSV should have isolation valves and a bypass globe valve and the MSV should be a control valve with a manual handle. System problems related to HTF freezing, UPS and solar tracker failure were observed. HTF is a critical component of the power plant and the system and the pump seals should be designed to properly accommodate for anti-freeze protection. The lesson learnt in the system problems is that a temperature control strategy should be designed for HTF freezing. It is important to ensure uninterrupted power supply to the plant for smooth execution of work and commissioning activity of the plant. The experience obtained from commissioning of the power plant is documented and available in public domain so that similar problems can be avoided in future. 6. Acknowledgment We are grateful to Ministry of New and Renewable Energy (MNRE), Government of India for funding the project. 7. References 1. Desai N.B., Bandyopadhyay S., Nayak J.K., Banerjee R., Kedare S.B.. Simulation of 1 MWe Solar Thermal Power Plant. ISES Solar World Congress, Cancun, Maxico, Nov. 3-7, 2013 (Accepted) 2. Desai N.B. and Bandyopadhyay S., Solar Thermal Power Plant Simulator, SOLARIS, Varanasi, India, Feb. 7-9, Solutia Inc., Therminol VP -1 Properties : St Louis- MO USA. 4. Leak Seal Experts : 449/B, Mundka, Delhi