Eshkol Power Station, Israel
EXERGY Examples for 2007, 2008 data at the Eshkol power station in Israel.
This document describes a number of EXERGY examples from the 228MW units (5-9) at Eshkol station in Israel.
There are long-term and short-term efficiency considerations. The former is called condition monitoring and is also of interest to the maintenance department. The first example (Figs 1 to 4) illustrates this around the November 2007 Unit 9 outage. Short-term analysis is illustrated in the section titled Efficiency Alarms below. It demonstrates real-time alarming and diagnostics of efficiency excursions. Values are recorded and plotted at 10 minutes interval. The examples can be reproduced on site.
The plots in figures 1 to 4 show various Unit 9 parameter trends starting from September 2007 till 21 December 2007, filtered for a load range of 220MW to 228MW. The gap in the plots symbolises the outage period. It is shown that the thermal performance of the unit is somewhat worse off after the outage and it is explained how this conclusion is derived.
Consider Fig.1. The black line of the top graph is the generator’s power output ranged on the y-axis. (If this graph is reproduced in Plant Explorer, then the y-axis display can be changed to another parameter by clicking on the desired tag on the right-hand side). The red line tagged 1900.LF, (hereafter loss), is the Unit loss as a fraction of the exergy of fuel. The blue line tagged 1900.1L is the station’s conventional unit heat-rate in Kcal/KWh. It can be seen that the heat- rate improves (is lower) after the outage whilst the losses are increasing: This is so because the heat-rate is not corrected for the sea temperature which is 10 degrees lower in December than in September; the loss (as well as the thermodynamic efficiency) is corrected in a natural manner and hence is the true indication of unit performance. The second graph indicates the turbine cycle loss (green line tagged 2800.LF) which remains unchanged (or even slightly decreases) from just before the outage to after the outage. The third graph indicates the boiler’s loss (heavy pink line tagged 2700.LF) as well as the boiler’s so called First Law or thermal efficiency (light pink line tagged 2700.1L). 2700.1L shows an improvement whilst 2700.LF shows deterioration; however it is the latter that is the truly full indicator of boiler performance because it lumps in the steam temperatures and pressures as well as the stack “loss”. Since the 2700.LF trend corresponds (correlates) to the post-outage Unit loss (1900.LF) increase, the boiler contains the root-cause of the unit’s worse-than-before thermal performance.
In fig. 2, a 4th graph is added to Fig.1 to show the thermodynamic efficiency (heavy pink tagged 1900.EF). It can be seen that it drops slightly from just before the outage to after it, as one would probably see from a properly corrected unit heat-rate.
In figure 3, and the ones that follow, it is demonstrated how the root-cause of the post-outage unit loss increase can be tracked down. Essentially one simply follows the hierarchical tree of the plant part of which is shown on the left-hand side of the figure. One easily notes that the boiler loss is due to the loss associated with the “steam and flue gas” subprocess (the Hebrew translation is wrong, should be flue gas rather than ignition gas) (green line tagged 3751.LF) which increased in tandem with the boiler. To see why this happened, look to Fig.4 below.
In figure 4, one can clearly see that it is the combustion process (on the graph, this is the red line tagged 5702.LF) that corresponds to the boiler loss shown on Fig.3 above which in turn corresponds to the Unit loss of Fig.1. If this graph is recreated in Plant Explorer, one can see what actually happened by clicking on  on the left-hand side and examining the associated measured values. Figure 4 shows the results. It is clear that the secondary air flow (second graph grey line tagged 206.FW) records a post-outage increase whist the temperature of the secondary air (third graph blue line tagged 207.TP) decreases - a well-known recipe for poor combustion.
The 2 figures below show a non-event, the trend of various components efficiencies of Eshkol ‘s unit 6. Fig 5 shows the trends for the first week of October 2007 of the superheater efficiency (5755.EF) in black, the reheaters efficiency (5757.EF) in red, and the generator’s electrical power (5702.PW) in blue. Note that for power variations of 50% said efficiencies vary by only 5%, hence the .EF are good condition-monitoring indicators.
In figure 6, the trend of the low pressure turbine efficiency was added (green line tagged as 4852.EF) with duration reduced to first 3 days of October 2007.
Finally one can add benchmarks to main trends which appear as dotted lines with the same colour. Said benchmarks encapsulate the efficiency as it was upon commissioning, adjusted to weather conditions, load, and fuel used. One notes that whilst the L.P turbine performs satisfactory, the superheater and reheater are off the mark.
Detection and alarming of efficiency excursions is a major feature of the EXERGY system which among others distinguishes it from conventional packages. The efficiency alarms identify deterioration of unit thermal efficiency (or specific fuel consumption) in real time, quantifies it in monetary terms, tracks down the root cause of the excursion. and notifies the Unit Operator (say). If a simple, costless remedy can be found, Exergy recommends it. Alarming can be implemented in two ways.
1. Alarms can be sent via email to a custom list of addresses.
2. An alarm client can be installed on any PC. This links to the server using internet protocols.
All alarms go to a database table. By default, this is in an MS Access database but it can be located within an enterprise database
The alarm client is a system tray component. An efficiency alarm is shown on the right.
The tray icon flashes and beeps once. Clicking on the tray icon brings up the monitor.
Screenshot of excursion alarm:
Screenshot of alarm analysis using a EXERGY Plant Explorer. Similar graphs can be viewed PI Processbook using an EXERGY-supplied template.
The example above exhibits a common efficiency excursion. Power output (in black) has been constant along the interval at 225 Mw yet the heatrate tagged 1900.1L increases . EXERGY has automatically traced this excursion to 3 processes the 2 air heaters (tagged 5703/LF and 5704.LF ) and the the combustion component (5702.LF, brown in lower graph). There are no other hints on this particular alarm , but the poor heatrate is probably associated with the secondary air temperatures . Either way an operator is aware of the actual event on the plant; Heatrate would be improved if operator
Took action associated with these 3 processes only.
The example below is a more recent similar event, except that the cost is by far higher, and but the gas temperature here went higher rather than lower. It appears that there is a systematic mistake in setting the air temperature that Operators repeats
All of the code used in this demo, assumed that Eshkol units have a configuration as drawn below: