Critical Gas Turbine Parameters

Hello everyone...

I am working on gas turbine operation (19.6 MW, frame-V, GE make turbine, Brush make Alternator). I would like to know what are the MINIMUM turbine parameters which are to be monitored by us to keep our turbine healthy (log sheet is too huge to monitor) along with its set points.

Thanks and regards
 
jagriti,

Which parameters to pay the most attention to ....?

Let's address the issue of the 'log sheet' first. The purpose of log sheets is to be able to spot a trend. A trend is an increase or decrease in the value of a parameter over time. The time period could be an hour, it could be several hours, it could be a day, it could be several days, it could be a week, it could be several weeks, it could be a month, it could be several months, it could be a year, or several years. But, the purpose of values in a log sheet (contrary to popular belief) is to be able to see (spot) a trend which is occurring.

For example, let's say one of the values in the spreadsheet is cooling water pump discharge pressure. And, let's say that when the cooling water system is in a new and clean condition the cooling water pump discharge pressure is approximately 4.5 barg. Over time, the cooling water (which usually includes some added liquid called coolant or even anti-freeze) and that can eventually begin to deteriorate and cause build-up of solids in various parts of the cooling water system, and it can even lead to increased wear on the cooling water centrifugal pump. If left unchecked for long enough, or if the wrong water was used to fill the system, these solids can begin to collect in the strainer which is usually upstream of the cooling water pump inlet (suction). This can eventually lead to a decrease in the flow--and pressure--of the cooling water pump discharge. Let's say on your unit the low cooling water pump discharge pressure was 2.0 barg. And, that the pressure had decreased to approximately 2.1 barg over time.

The decreasing cooling water pump discharge pressure values in the log sheet should alert an observant and conscientious operator to a problem with the cooling water system. As the problem gets worse, the cooling water pump discharge pressure will continue to decrease, and the rate of decrease could increase also. Log sheet values are intended to help spot possible problems with unit and/or auxiliary operation, to prevent a problem before it becomes a serious problem affecting unit operation. If the cooling water pump suction strainer gets too full (dirty) the flow of cooling water to the L.O. coolers will be too low to properly cool the L.O. flow to the bearings, and the bearing metal temperatures and bearing drain temperatures will be seen to be increasing (from the values entered into the log sheet). If nothing is done, eventually the High L.O. Temperature Alarm will be annunciated by the turbine control system, and at that point the operator will need to make a decision: to lower load and/or shut down the unit to determine why the L.O. temperature has gotten too high. (By the way, for GE-design heavy duty gas turbines, the normal L.O. temperature leaving the L.O. coolers should be approximately 51-54 deg Celsius. And the High Alarm setpoint is approximately 74 deg C. So, an attentive operator should have spotted the problem of increasing L.O. temperatures in the log sheet LONG before the high alarm setpoint was reached!)

BUT, had the operator (or the supervisor reviewing the log sheet values) spotted the increasing L.O. temperature over time AND the decreasing cooling water pump discharge pressure over time the alarm and potential need to reduce load or shut the machine down to resolve the problem could have been avoided. (There is usually a manual "blow-down" valve on the cooling water pump suction strainer, which can be used to periodically remove some contaminants and solids from the strainer while the system is in operation. This could have been used to try to help restore the cooling water pump discharge pressure long before it dropped precipitously low, and by observing the liquid and solids exiting the blow-down line the cause of the contamination could be investigated.)

Also, the decreasing cooling water pump discharge pressure should have also alerted an observant person reviewing the log sheet values to a problem which would have not resulted in increasing L.O. temperature.

Log sheets are a KEY part of properly operating any piece of equipment. The value of a parameter at any given point in time is not really important (unless it is excessively high or low), but the trend of the value (increasing or decreasing; increasing quickly; decreasing quickly) is what is important. And, the time period is also important. If log sheet values are only entered once per hour, the change in value in one hour might not seem important (such as the cooling water pump discharge pressure decreasing, slowly, over several months, or even years; and the L.O. temperature increasing slowly over several months.) But, taken over time the increasing or decreasing values ARE important and can portend possible problems before they actually become problems.

EVERY value is important, but more important is the value over time--the "trend" of the value to be increasing or decreasing, and how quickly it is increasing or decreasing. Many people ask, "Is 3.9 barg a good value of cooling water pump discharge pressure?" That's impossible to say for certain. The proper response to that question is, "What was the value of cooling water pump discharge pressure one hour prior, four hours prior, one day prior, two days prior, a week prior, a month prior, two months prior, six months prior, after the last maintenance outage?" Because the instantaneous value of cooling water pump discharge pressure--unless it is excessively high or low--is not really important. It's how the value has been changing, if at all, over time. And how fast it has been changing, if at all, over time. If the cooling water pump discharge pressure has remained in a range of 3.9 to 4.3 barg for months or even years, then 3.9 is a "good" value.

Trends can be used to spot a problem BEFORE an alarm or shutdown or trip is annunciated or initiated by the turbine control system. An alert and observant and conscientious operator will always be on the alert for trends in values. On starting his shift, and operator should be briefed by the operator going off-shift of any changes in operating parameters. And, the on-coming operator should then be looking at the log sheets and the HMI to see if any of the values he (or she) normally looks at have changed by much since his (or her) last shift ended. And, by looking at the log sheet values over time a trend can be spotted and the supervisor can be alerted to a potential problem BEFORE it becomes a sudden and real problem which might result in a shutdown (automatically initiated by the turbine control system, or one which is manually initiated by an operator in response to an excessive parameter value), or a trip of the unit.

When looking at log sheet parameter values, one also needs to try to compare values from similar operating conditions. That means that one shouldn't necessarily try to compare CPD (Compressor Pressure-Discharge) at 5% load to CPD at 100% load. One should be comparing CPD at 5% load to previous values of CPD at approximately 5% load, and CPD at 100% load to previous values of CPD at 100% load. And for CPD, the ambient temperature of the air entering the axial compressor of the turbine is also important. One can't really compare the CPD at 100% load and 12 deg C ambient temperature to CPD at 100% load and an ambient temperature of 40 deg C. (Ambient air temperature can greatly affect CPD!)

In my estimation, the most important parameters--over time--are exhaust temperature spreads and CPD and fuel flow rate and fuel pressures (particularly for unit burning gaseous fuel(s)). While liquid fuel supply pressure is also important for a turbine burning liquid fuel(s), it's not as important as gas fuel supply pressure and what's called P2 Pressure (the pressure between the Stop-Ratio Valve and the Gas Control Valve(s)).

Exhaust temperature spreads can indicate problems with plugged fuel nozzle orifices, or if the unit is burning liquid fuel it can indicate a problem with liquid fuel check valves, and purge air check valves (if the unit also burns gas fuel(s)). And, the trend of exhaust temperature spreads can indicate problems long before they become actual problems, which can result in an automatic turbine trip by the turbine control system. Exhaust temperature spreads can indicate problems with cracked combustion liners, or crack combustion transition pieces, or broken or breaking hula skirt seals or nozzle side seals. Some of these conditions can occur very suddenly, but things like plugged fuel nozzles or failed or failing check valves can occur very slowly, and be spotted by reviewing and analyzing exhaust temperature spreads.

CPD can indicate a dirty compressor, and/or dirty axial compressor/turbine inlet air filters. A dirty compressor decreases unit efficiency, meaning it will require more fuel to produce the same load as when the compressor is clean.

If the generator has inlet air filters, dirty generator inlet air filters will cause the generator stator temperatures to start increasing over time.

Yes; yes; yes; for MOST critical parameters the turbine control system will "protect" the turbine and shut down or trip the turbine to protect it before any serious damage might occur. BUT, by spotting trends using the log sheets a good operations department--and even a good maintenance department--can be proactive, meaning they can take preventive action BEFORE a shutdown or trip can occur. They can schedule an short shutdown to investigate or resolve a problem BEFORE the turbine control system initiates a shutdown or trip. It's much better to be proactive than reactive (reacting to alarms and shutdowns and trips!).

I often find that when reviewing log sheets AFTER a unit trip, that trends can be seen to be developing over time which, if they had been spotted by alert and observant and conscientious operators could have prevented the trip, and the loss of generation and revenue. Many sites do not look at log sheets after a trip, and they simply re-start the turbine which eventually results in a subsequent trip, or even multiple trips--which could have been avoided had the log sheets been properly reviews and analyzed.

Many operators think log sheets are a bother; and that they were invented by supervisors and managers to keep operators busy during the "boring" periods of operation on a shift. <b>NOT SO!</b> Log sheets are a critical part of operations and maintenance! Failure to properly review log sheet values can lead to lots of unintended shutdowns and/or trips, and trips from load can cause damage to hot gas path parts, so they are to be avoided whenever possible (by being proactive and taking appropriate actions such as unloading or shutting down the unit to resolve a problem before it results in a trip).

Maintaining a "healthy" unit is also achieved in a VERY LARGE PART by using log sheets to spot trends. A dirty compressor can result in loss of load and decreased efficiency (more fuel for the same load as when the compressor was clean). So, CPD is a very important factor to monitor for trends. All compressors get dirty over time, and there is usually a point at which the plant management will shut the unit down and perform an off-line compressor water wash to clean the machine and restore lost power and performance. That can be seen most by looking at the trend of CPD over time. CPD will decrease as the compressor gets dirty (becomes "fouled") and after the off-line compressor water wash there should be (will be--if a proper off-line compressor water wash was performed!) a marked increase in CPD (at Base Load)--and an increase in electrical power being produced by the generator (at Base Load). And, if the fuel flow-rate is being monitored, the fuel flow-rate should decrease slightly but markedly (at Base Load) after a proper off-line compressor water wash has been performed--because the unit will be more efficient, requiring less fuel for the same load, or producing more load with the less fuel than when the compressor was dirty/fouled.

Exhaust temperature spreads and CPD; those are the most important <i><b>turbine</b></i> parameters to be observing. As the unit has many auxiliaries, the are many other parameters to also be observed. And the best way to control and protect a turbine--and to keep it "healthy"--is to properly review log sheets. Is it easy (to review log sheets)? No. Is it fun? No. Is it time-consuming? Yes. Because some parameters have to be compared to values last week and last month and last year. Is it important? <b>YES!!!</b>

Bearing metal temperatures, and bearing drain temperatures, and bearing vibration readings are also important--again, not just the instantaneous values, but the trends of the values over time. By reviewing the log sheets every shift, or twice a shift, or more, trends of bearing problems can be spotted increasing quickly and supervision and management can be alerted to prevent a trip or catastrophic failure. And, if you're the one continuously spotting trends and helping to reduce shutdowns and trips, well, that's good for you, isn't it?

Every site has usually had multiple issues with one or more auxiliaries or pieces of equipment. So, the operators who have been operating there for a while will concentrate mostly on those auxiliaries or associated parameters. They will tell you (or they should tell you) what those pieces of equipment associated parameters are.

But, log sheets--no matter how long they are--are critical to proper operation and maintenance of the unit and its auxiliaries. Again, just looking at a value on an HMI screen doesn't indicate a potential problem. It's how has that value been changing over time, and how quickly (or, slowly) has that value been changing over time. Many values will only change slowly, or should only change slowly. Many other values will change quickly, or if they change quickly sometimes it's impossible to respond to in time to avoid a shutdown or trip. But, most problems can be spotted by watching trends.

Hope this helps! Operating a GE-design heavy duty gas turbine is much more than just silencing the audible alarm horn. And, much more than just sitting at the operator's desk (or console) and looking at screens. It's about being proactive and spotting trends (when possible--again, some problems occur very quickly and don't allow for operator intervention). AND, it's about Alarm Management. It's about being aware. It includes understanding the systems and components of the turbine and how they work together to efficiently produce power. It's not just about sitting at a desk/console and silencing the audible alarm horn every time it sounds (which for some Mark VIe turbine control systems can be VERY often!). Operations--proper operations--is about monitoring parameter values for trends, responding to alarms properly (called Alarm Management), and being proactive. That all happens when operators are properly trained, have some experience (the more good experience the better!), and understand how the turbine and auxiliaries should operate--so they can tell when they are NOT operating properly and can contribute to solving problems quickly and keeping the unit running and reliable. A reliable unit is just as important as a healthy unit!

Lastly, you're new here (and we can always spot the newbies--they have LOTS of questions they want answered in a very short period of time). That's understandable, and commendable. The GE Speedtronic/Mark* turbine controls community has been active here on control.com for almost two decades now. In that time, a LOT of issues have been covered--sometimes repeatedly. And a LOT of questions have been asked and answered. The beauty of World Wide Web forums like control.com is that past issues and questions (called "threads") are archived (stored) and can be accessed for many years after they were initially opened or started. This means there is a LOT of information (a WEALTH of information!) available to anyone who uses the 'Search' feature of control.com, which is cleverly hidden at the far right of the Menu bar of every control.com desktop version webpage. (The Search feature of the mobile version of control.com is accessible under the Control.com tab.) It is recommended you use the Search Help the first few times you use the Search feature; the syntax of the search terms is not like most present-day search engines--but it's just as fast and powerful! You will be quickly become accustomed to the syntax the more you use it, and you should use it because, again, MANY of your questions (the ones you have now and the ones you're going to have!) have already been asked and answered.

And, the REALLY nice thing about control.com readers and contributors is that they usually take the time to respond to the information provided to them so that others who read the threads can see if the information was helpful, or not. (Hint. HINT!) May people also click on the 'Thumbs Up' or 'Thumbs Down' icons at the top of every thread entry to indicate their level of satisfaction with the information or question. (We would like if people who click on the 'Thumbs Down' icon to write to say what part of the response they didn't like.)
 
Hello CSA ....

Excellent write up for beginners like us and it is explained so nicely. Thank you sir and must say I am overwhelmed...This will definitely help us to for better understanding.

Regards
 
Top