Home
Products
Applications
Parts & Service
Rental
News & Events
About The Company

Load Bank Testing of Generator Sets

by ‎10-04-2011 11:03 AM - edited ‎10-04-2011 11:04 AM

pp-benz-banner.jpg

Generator sets represent a sizable capital investment, and are seldom installed as amenities of convenience. Their failure to perform properly when called into action is almost always accompanied by potentially dangerous conditions and/or risk of significant financial loss. Periodic load bank testing is an important element of any comprehensive preventive/predictive maintenance program to ensure the genset is capable of accepting certain building loads when required. But, load bank testing is seldom specifically required by local codes, unless of course, the genset operates as an emergency standby unit required by building safety code. These gensets are regulated by applicable NFPA 70, 99, 110 (Standard for Emergency and Standby Power Systems), and The Joint Commission standards.
 

  • Facility managers or plant engineers with responsibilities for genset maintenance are well aware of the need for preventive/predictive maintenance and periodic operational test-running of their power generation system, but, is the same consideration given to load bank tests?
     
  • Is it common for facility managers or plant engineers to include load bank tests in their preventive/predictive maintenance programs if the facility is brand new, or is this something that is considered after the fact, maybe 3-5 years after the facility is commissioned and operational?
     
  • How often do you conduct load bank tests at your facility?
     
  • Do you run load bank tests concurrent with other maintenance testing? If yes, what types?
     
  • Do your personnel perform the testing or is it something that is contracted out?
     

Please post below to share your perspective.
 

Comments
by on ‎10-06-2011 03:11 AM

In the UK some few sites will have permanently installed load frame test equipmet, the others we always recomend at least annual testing.

Issues with wet stacking are a constant reportable problem from the end users, and we would always recomend load frame test prior to engine clean for this.

An issue we have here is site operators understanding of what the equipment can and cannot do, they may have recieved advice from some "consultant" who insists on 110% test when the unit is sold as pure stand-by rating.

With the complexity of modern buildings and the favorite pastime of installing all generation equipment either on the roof or under ground, unless some thought has gone into the installation process a load frame connection point may be totally inaccessable so load testing may never take place due to vast cost for the connection.

 

We have several multi-engine sites that have never been tested due to these costs and the endusers reluctance to allow any testing in case of a mains failure, the actuall mains fail testing is as important if not more than the load frame in my opinion but both should be combined to give indication of system operation.

 

When load frame tests are carried out we get the equipment installed by contractors and we will then carry out a four hour test at rated load, with warm block load tests at the end.

by muralcr on ‎10-06-2011 06:58 AM

i am happy you mention about load bank testing of generator at customer end ,in india once the set is installed no body bothers about load test on load bank.its amusing even the delears talk in terms of fuel consumption in Ltrs/Unit.if i remember right only on board ships i was testing the engine on load banks once ain two years, in Indian Navy.

by on ‎10-06-2011 08:01 AM

Before you recommend repetitive portable load bank testing, you need to ensure the application is designed for repetitive breeches. The biggest problem we deal with is the integrity of molded case breakers using aluminum mechanical lugs. These lugs are not designed for repetitive manipulation, even when paste lube is used.

 

If you know that a particular generator set will need maintenance level load bank testing due to light building loads or code compliance, a quick-disconnect cam-lock box or similar bolt bar box should be utilized.

by on ‎10-06-2011 08:18 AM

Difinetly load bank testing is critical to  ensure the responsibilty to our product. this is absolutely necessary to do things the right way, but the lack of responsibility is the reason why we suffer the consecuences of our actions.

 

Great article.

by rgarcia22 on ‎10-06-2011 10:28 AM

In more than one occassion we have been offered resistor banks test for gen-sets installed in mines before commissioning. In our experience these tests are rarely carried out due the load bank transportation cost and the test cost itself. Besides that the thermal load, how can you test the electrical performance of the gen-set (the test load offered was purely resistive, so the power factor could not be changed)?

by Eelizalde on ‎10-06-2011 10:32 AM
Load bank testing is important but doing it properly is even more important. Some experts say that partial load testing at 75% is adequate, but in my opinion, 100% load testing is a requirement. Verifying coolant temperature and oil pressure, etc. are important but I also recommend that an infrared gun be used to check and record temperature at electrical terminations. Checking biatteries and block heaters should also be done. After all , most generator failure occurs because of battery failure.
by on ‎10-06-2011 11:40 AM

Routine Load Bank Testing of all generators in a stand-by application can be a critical component of Healthcare acredidation. All Healthcare facilities who are required to adhear to NFPA 110 or JCAHO requirements and standards, must conduct Load Bank Testing annually if they don't meet the minimum requirements of NFPA and every three years regardless if following JCAHO requirements.

by DIKE on ‎10-06-2011 12:08 PM

On a more serious note ,load bank testing is very important  on the customer end so as to save the life of the generator from crashing out early more especially in my own part of the world where a lot of factors can contribute to the poor performance of the gen sets.Above all a standard must be set to ensure strict compliance.

 

Dike Kingsley

Nigeria

by lcampbell on ‎10-06-2011 01:52 PM

my predecessor would say not required, because we have numerous outages every year, and he thought as long as all goes smooth during those outages, gen sets are fine. on my watch, the last 4 years have been load test years, and many problems surfaced when putting 100+% load to them individually. fires, cooling problems, fuel problems, governor problems, etc. all show their ugly head when pouring the coals to them. problems that do not show themselves with 60-70% loading, and problems i do not want when they are asked to back up our loads. i have 9 mission criticle gen sets, and i load bank them every year. that way i get to worry about something else.

by markslum on ‎10-06-2011 04:29 PM

We generally recommend our customers peform at least a 75% load test once per year.  We have programs available too that allow the customers to get paid by the local utility for going "off grid"... Get paid to get load tested?  Great deal IMO.

 

Mark Lum | Vice President & Principal

Worldwide Power Products

5901 Thomas Rd. | Houston, TX 77041

(o) 713.434.2300 | (m) 713.264.2030

 

View our current inventory:

http://www.wpowerproducts.com

by Kengine7 on ‎10-07-2011 03:00 AM
With all due respect to the community, does any historical data support regular load banking? I am searching for a reason to load test. The following statement is made up, but this supports an action to change or add to a preventive or predicted maintenance routine: "Due to the number of flywheel UPS systems failing to provide seamless transitional power between utility failure and facility owned emergency power, all such systems shall be routinely tested under load. Regular testing of flywheel systems under predicted load conditions will likely provide equipment owners data that can be used to mitigate future UPS failure and more importantly, assure end power users' reliability as close to 100% as reasonably achievable." Again, the flywheel UPS statement is not true, rather, illustrative of a cause to load test the equipment. My assumption is that the stress upon a system which is routinely not in service, loaded, shall expose weakness in the system under test. The test serves as a means to acquire information which can be compared with historical failures. All similarities between known failed systems and tested systems will provoke change. The first response will be local repair or part exchange, while engineered changes will follow resulting in design change. Years of this process may yield new maintenance procedures which exclude load testing. This is the engineering model as I imagine it. Standby and backup power systems are continually improving and have been in existence for a very long time. May I suggest that except for emergency use as provided by administrative procedure by political and industry authority, load bank testing after manufacture testing does little more than provoke failure, waste fuel, and decrease equipment value. I will gladly resign my position on load testing in exchange for clear regionally equivalent data showing load testing prevents backup and standby power failure. Note the following assumptions: equipment in question is under professional supervision; equipment is installed new or removed and reinstalled by original purchaser; routine run testing is performed by regular assigned personnell; equipment is known to be manufactured under industry quality standards; equipment is installed as required by manufacture and NFPA or other building code applicable. Simply, quality equipment, installed per all requirements, should be expected to perform under non-emergency conditions when needed following a regular history of maintenance and exercise. If information exists that shows fast starting and voltage/speed regulation at proper unloaded coolant temp, oil pressure, other normal indicators is not a reliable predictor of loaded performance, it would then be the basis to the authors questions.
by on ‎10-07-2011 08:34 PM

I'll take a crack at responding to this. Based on 30 years of experience, not only with CAT but many other manufacturers units as well, I can say quite comfortably that regular load testing is beneficial to the unit, and improves reliability. Please note the bulk of my experience is from southern California, but as a contract startup and service engineer to CAT for several years I worked on systems all over the USA and the world and found my experiences to be similar no matter where I was.

 

Typical generation units using industrial diesel engines (we could go into other prime movers but the bulk of my standby experience is with diesel powered units) typically have engine support systems designed for maximum operation at maximum design limits. Lets use a radiator cooled engine cooling system as the first example.  At rated speed no load the engine cooling load on the radiator is less than 10% of the maximum heat rejection of the engine.  If you test at no load, even with a partially fouled water or air side, you will still maintain an acceptable engine jacket water temperature.  Remember, engine cooling regulators (or thermostats) operate to maintain a minimum engine temperature, usually by bypassing cooling water from the engine outlet back to the engine inlet.  Once the regulators are fully open, engine temperature is controlled by the ability of the cooling radiator to remove heat.  So if you are running the engine, observing only the displayed engine jacket water outlet temperature, how, at a very small percentage of the system heat load, can you assume the cooling system is working as designed based on on temperature? Now if you went to a more comprehensive monitoring of the engine coolant temperatures, both leaving and entering the engine, and compared to data from engine startup with new equipment to current readings, you might be able to predict a loss of cooling system efficiency.  But would you take the time or spend the money?

 

As an example, a world wide freight company with a major hub operations terminal at a southern California airport recently suffered not one but both units failing after a short time on-line. The units were over 10 years old, engaged in a regular contract that by the way included an annual load bank test, regular fluids monitoring, and careful oversight not only by the local dealer (who was well experienced  in taking care of standby units), but by company equipment and assets managers who manage fleets of on hiway trucks and scores of other equipment daily, with great success according to the industry and their peers.  The local utility asked the site to go on generators during an extended heat wave relieve the grid load, the site operators felt confident their generators would handle the load ok since they were well maintained and tested, at just over 35 minutes into the event, the first unit failed on engine coolant overheat, and the second unit was in high cooling water temp alarm. The load shed system functioned as expected and the second unit failed about 8-10 minutes later, also on engine coolant overheat. So why had a well designed well maintained system failed when called to operate during an event? And to add the average load on each unit was about 85% of rated. A couple of things contributed.  The design used 95 degrees F as the maximum ambient, the ambient the day it failed was over 100 degrees F.  The coolant was regularly changed every two years, with annual sampling performed by both a CAT and independent lab, with no issues reported. But when the radiator was inspected, air side fouling and water side corrosion were both evident, neither would be considered bad by itself, however in combination with minor fouling on both sides, and ambient temps in excess of design it was enough to cause a failure.  We also found one of the high jacket water alarm switches was not functioning. So after we found all this out we stepped back and reviewed the maintenance.  First we found that all load bank testing was performed in January, the coolest time of the year for this site. We also found that load bank testing was performed at 80% of rated load, instead of full load, because bringing in the right size load bank was considered an "excessive expense" since the unit loads were only about 85%. As we dug a little further we found the previous years load bank test had been cancelled, after all, is was a well maintained unit, not that old, and testing every year was seen as a waste of fuel, resources and a negative effect on the environment (paraphrased from an owners document). After the radiators were removed and cleaned, we performed a test of the system, at 100% load (both real/reactive power).  We confirmed that the radiators would in fact handle the site load of 85% per unit at an ambient of 110 degrees F by performing an ambient capability test, although the units would be at the alarm setpoint of 215 degrees F.  During the load testing at 100% per each unit, and load sharing at 50% per unit we also found some other issues.  The fuel supply pressure to the engines was at minimum above 80% load, as we further investigated we found that the fuel supply lines were deteriorating and their inner liners were starting to peel away at higher fuel flows causing a restriction. From 0-50% load the fuel pressure was near maximum and no problems were thought to exist, above 50% to 80% the fuel pressure slowly dropped but was still in the "good" range, as we got towards 100% the engine frequency fell off and we had to make a repair to continue. We also found a problem with one of the generators, the field voltage at 50% kVA load was near maximum rated field volts, at 100% the regulator was just below its maximum output, although the voltage was at rated and there didn't appear to be any problem. It turned out it had a bad diode in the rotating rectifier assembly.  The list went on, and is still generating a fair amount of grief for the dealer who maintained it and the owners people who were in charge, as the total price tag for the outage was in the millions of dollars ( the issue is still in litigation, sorry for the lack of more details).

 

Now I'd like to say this was the exception, in general modern standby generators driven by diesel engines have a high degree of reliability, and frankly can withstand a fair amount of abuse for a relatively long amount of time.  But new design engines, designed to meet increasingly strict emissions standards actually are less tolerant to light and no load operation than older engines. Also newer engines that have high power densities and are highly turbocharged are prone to overfueling and poor combustion at low loads, so when they run, byproducts of incomplete combustion contribute to "wet stacking" and "slobber". Unless the engine is run at a load high enough to generate high exhaust temperatures and flow velocities, the deposits will coat engine exhaust surfaces including turbo turbine wheels, exhaust piping and mufflers/silencers (and in mine and increasingly other parts of the world, installed exhaust aftertreatment systems).  Initially this is only a nuisance, however allowed to build up, it will affect engine performance (turbo doesn't spin up like it should because of all the gunk on the hot wheel), and allowed to go on further will generate a fire hazard buy building up combustible debris in the exhaust system. So if you think it's a waste of money to regularly load test, see how much you'll spend after a muffler fire on a unit in a downtown area where the minimum fire response is four companies, along with police and EMS. Funny thing is it stills happens around here almost every year.

 

How about your on site fuel?  Will you turn it, use it in something else (better make sure you paid the right taxes on it), or leave it in the tank and hope it will be ok when you really need it? While there are a number of options, your overall fuel storage and supply system design will dictate it actual maintenance needs. A larger system with storage tanks, day tanks and transfer pumps all need regular testing to assure they'll work properly when needed.

 

I quite frequently get around system that have been not tested for some time, either due to financial issues, like a building going thru multiple management/ownership changes and the little standby in the basement gets overlooked., or opinions such as the ones you expressed above, that regular testing contributes to more than problems than it prevents, especially now that "green thinking" is becoming more prevalent. In just about every case I know of, not doing the testing, or improperly doing the testing contributes to a reduction in overall unit and system reliability, because if you're not testing the generator unit, you're also not testing the ATS, emergency power distribution, controls, alarms and support systems. I should also add while the bulk of my experience was with a CAT dealer for 27 years, I am no longer in the business of selling generators or maintenance contracts, I do do power system evaluations troubleshooting and repairs, but if you don't regularly properly test and maintain your standby power system, I actually make more money, because when it doesn't work, you're in the dark, your line is down, you send employees home or can't answer the phone to take care of your customers.

 

One more thing that I'm seeing as an increasing trend. When I started in this business, I found many of the commercial insurance companies, like Hartford and others were very proactive in getting their clients to perform maintenance and testing to reduce claims.  Now days I'm seeing the insurance folks not being involved, and quite happy to come in after that fact and ask if you've been doing your due diligence, maintaining your equipment to industry standards and performing the required testing, and if you say no, guess what, they don't want to pay, sometimes not at all.

 

Sorry for the long winded reply, this is a complex subject and I'm glad you posed the question. It is something I deal with regularly and lately find that many of the manufacturers and dealers aren't doing a good job educating their customers on what the real benefits are to regular testing and the drawbacks of not doing it.  And yes, there are some things that could be done differently. I'd prefer to see units load tested against the grid when possible instead of wasting the energy, but in units that stand alone testing with a load bank or actual site loads is needed to assure it does what it's supposed to. A good example of that is I just got done reviewing a site where the dealer/contactor did a load bank test on a new unit, the customer had not defined performance criteria in the sales contract, except to say it need to accept a 50% step load, so the unit was tested and it "passed", everything got wrapped up, all were happy until a recent outage, the new unit dropped the PDU's in the data center. Seems they had a voltage and frequency tolerance of 5%, not the 16% voltage and 7% frequency the unit actually dropped when tested at 50% load.

 

We've had other discussions on this site for dealing with things like fuel, starting batteries and lube oil.  Not really any easy answers, but I will pose two questions to you in closing.

 

If no load testing provides a reasonable assurance of loaded engine performance, why do many states now require the use of a chassis dyno to emissions test cars?

 

If your resting heart rate provided a good indicator of your overall health, why would your doctor order a stress/treadmill test?

 

I hope this at least partially answered your questions, I look forward to any comments or further questions you may have.

 

Mike L.

by Kengine7 on ‎10-08-2011 04:53 AM
Thanks, Mike. That's exactly what I wanted: historical evidence from real events. So you have thirty years of facts supporting load testing; absolute problems that were unknown because load testing was not performed-and you represent a small percentage of all similar events and evidently an employee of a very large manufacturer of power generation equipment, so why does this blog exist? The manufacturers have the responsibility to document facts because they HAVE ALL the data. I remember when the speed limit was reduced to 55 and Detroit Diesel informed Greyhound bus line to expect significant increased fuel costs because the torque/RPM characteristics were optimal at 70 for all installed engines to date. No debate about it: GM had the facts and passed them on. Seriously, I resign my opinion on load testing. You make an excellent argument although the bad diode at 100% load at just under full field, is intriguing... The anecdotes about emission tests and treadmill tests are out of context in my opinion. This blog would not exist if the equipment was contiuous use.
by mkziawan on ‎10-08-2011 05:57 AM

I have here marine vesion engines and other power generators in Pakistan .Usually the peventive and corrective maintenance are being carried out regularly but only marine engine load test facility is available as load testing van and load test is being carried out regulary after one year at 100 ,105 & 110% repectively according to the date of installation and the age of the engine and Generator but here Generator used for power Generation rating 50 Hz frequency are being tested up 80% or less due to nonavailability of load testing faclility and have degraded their efficiency. I think it is very essential to load test of the Generator to up keep their efficiency and also main failure of the engine which can cause of severe damange and loss of major components of the Generator.

I can also decleare here that the other accessories are equally important  for testing periodically and the awairness in this regard is also required to up keepkeep the efficiency of the machinery equipments. for the future references manufacturer of the power generation  equipments are to educate the dealors to emphsised on the preventive maintenance regading the load testing as it is very essential as longer life of their machinery. The efficiency of the genset also dependable to the way of installation and frequent atmospheic changes  in area. Specially the genset fitted at high altitude have degraded their efficiency rapidly due to the moistuer and less availability of air flow. Simmilarly the gensets used for only standby purposes are found to be much problematic then the Genset used on regular basis. Load banking is also well known for the genset those are not used on regular basis and only operated when power failure occurres. I think as a maintainer of the machinery I have sort out that if you are efficient in dealing with prreventive maintenance then you have lot of time to awair of your machinery problems and before the time you have to approach your reasons of failure and also you have an efficient machinery in your hand all the time and you can safe your money which is used in bulk and also you have panic for repair work. It can also be known through the example here that the people have their regular visit to their medical practitioner and get their several test regularly ,they have in better health condition other then those might visit when they are felt in bad condition and visited the clinic and they have lot of time to recover their health.

ZAFAR IQBAL

Marine Engineering Officer

PNS Hafeez

Islamabad

by on ‎10-08-2011 01:50 PM

Kengine7,

 

As I stated in my original post, I get involved in this discussion on a pretty regular basis these days. I currently work for myself, I don't sell equipment or do maintenance contracts, but I do sell my expertise based on many years experience of doing both and some other associated services.  Manufacturers of gensets generally do a fairly good job of recommending maintenance practises, but they are broad based and don't always cover specific needs of a particular customer. Hopefully the manufacturers distributor/dealer picks up the slack and uses their experience to meet their local customers needs. Unfortunately not all dealers have a population of power generation equipment that they can draw experience from, and sometimes the guy recommending maintenance on your generator is way more familiar with farm tractors or irrigation pumps. But you may be disappointed to find out the manufacuturer doesn't have all the data. Your Greyhound story is a good point, DDA built the engines, Greyhound spec'd the transmissions, axels and tires, and managed their fleet to get maximum fuel efficiency and maintain schedules. DDA had very little idea how going from 70 down to 55 was going to affect most of their population, although since Greyhound was a large user they likely had some feedback from them. But it wasn't the manufacturer who had the hard data, it was the operator of a large and well managed fleet who did.

 

A blog and forum like this exists (at least in my opinion) to draw experience from a larger pool than what is typically available to a manufacturer, most use warranty information to define success of a product, and at the end of the warranty period, maybe some level of parts sales.  Some engage thru their dealers long term maintenance programs, mostly for prime power equipment, in which they also gain some knowledge.  But in reality, most of the valuable information regarding the long term health of standby equipment comes from the people who regularly work on that equipment and those who use it, and these are typically folks that the big corporations don't have much of an ear for. This blog is an excellent example of how a manufacturer can get real world info, it was targeted to users, trying to get their direct feedback in a forum that doesn't expose them to sales hype or a pitch for some miracal product to magically save them money, be more friendly to the environment or make them a corporate hero.

 

As for my anecdotes, they serve me quite well in my area, as many times I deal with people who have high positions in their companies, are extremely well educated and smart, and have no idea how some mechanical works.  The one for car emissions particularly hits home in California, as it caused the cost of a regularly state required test to go up, and reduced the number of places to get a test, there is also a large amount of hard data that supports the change in testing and it's positive impact on the environment.  The one about the heart test actually came from a customer, a CEO of a company who operates a large number of co-location data centers who had a rash of power problems, while trying to make the same points as I did for your questions, he suddenly told me that he, at 55 years old, had virtually never been sick a day in his life,and while he no longer had engaged in physical activity for many years he still felt trim and in good health, that his doctor flet it was time to have him take a treadmill test, no reason other than that his lifestyle could contribute to an underlying problem and he wanted to be sure he was as healthy as he thought, to cut this short he was in for bypass surgery in two days.  Needless to say, my discussion with him got much easier once he made that connection, and I have shared that comparison a number of times with other customers who just weren't quite getting it.  So I apologize if you found the reference out of place, but I have no idea of your postion, experience or background.

 

While I don't always agree with some things in this forum, I have found it a great place to share information and hopefully CAT actually uses this information to improves their products and their customer service, I do know several dealers use this as a place to get real world information and use what they learn to improve their operations. I also know several competitors who also log in or visit here to see what's going on here. And there are cetainly a large number of users who come here to get help when their normal avenues of service and support aren't getting them taken care of.

 

Mike L.

by Kengine7 on ‎10-09-2011 09:32 AM
I have designed several reactive load banks. Reactive loading is as close to actual loading a generator will see. I also support resistive load bank testing and frequently advise owners to do so. Under load I have unearthed the worst failures such as grounded field conductors, run away governors, ruptured head gaskets at coolant galleys, and often below nominal output power capability. It's important to network within our community. A huge fan of Ben Franklin, he would definitely support this exchange just like his Junto, 'who among us learned something that we can benefit in health, wealth, efficiency, safety, etc.' I take nothing for granted and listen to good council. Additionally, I am not at odds with you. I asked for facts supporting something I already do but routinely adds repair costs which is disturbing because the test evidently beccame the cause. I have also found that the best resource for help is usually the one with the least resources to help. A machine built to do a job and serviced as advised by manufacturer' data, should be ready to do the Job. These days opinion yields to fact. Thanks again for your real world experience.
by on ‎10-10-2011 10:24 AM

I like the exchange in this BLOG.  Glad to see everyone participating.  I'd like to add my two cents. 

 

The need for load bank testing seems fairly intuitive to me, especially when you consider the maintenance aspect of power systems.  As you all know there are small tweaks made on a fairly regular basis.  This could include the replacement of something like a fuse or filter to something much larger and more invasive.   I advocate load bank testing any time an unknown is introduced into the system. 

 

I'll use my UPS experience as an example.  Just because the technician tells me he has put a new board in my UPS, I'm not placing my mission critical load on that board until it's been tested under load in the environment where it will now live.  What if the revision level of the board was wrong?  What if the board suffered an ESD event during shipping?  Too many variables left to chance.  Put the system on a load bank and show me that all has been restored to working order.

by on ‎10-10-2011 11:27 AM

Kengine7,

 

You stated you were searching for a reason to load test, but later stated that your assumption is that the stress (brought on by load testing) upon a system which is routinely not in service, loaded, shall expose weakness in the system under test.  You also stated that "load bank testing after manufacture testing does little more than provoke failure". You already had the data you were searching for, right?  Real world experience with load bank testing.

 

We don't have all the answers, and we certainly don't have data showing load testing will prevent standby power failures, but as MikeL points out, load bank testing can identify potential issues before they become big problems at the worst time possible.  We're listening and appreciate you sharing your knowledge and comments with the community.  Together, we're building a better tomorrow, and maybe, a better PM program. 

 

Kevin

by tariq32120034 on ‎10-11-2011 12:06 PM

in Pakistan customer site genset and power houses there is no such practice like test on load bank but as we be a rental genset provider we always do load testing every before delivering the genset.

by FarangZoli on ‎02-21-2012 04:30 PM
Dear Cumminity Members, Please, advise me about the following: We are in the process to finalise our electrical power supply design to a Data Centre to be built in Brisbane, Australia. We have a normal HV supply to site via 11/0.4kV 1MVA transformer, which is backed up with an automatically started 1MVA CAT emergency standby generator with automatic transfer switch (ATS) employed at the Main 400V Switchboard. For critical data equipment supply we have also a redundant 500kVA industrial type UPS system with 30 minutes battery backup. My question is related to the regular generator testing with load banks. I specified load bank testing at relular intervals and the permanent load bank to be part of skid design. Is it practical or a mobile load bank would be a better option? Do you prefer the Community to use actual building load for generator testing instead of load bank testing? Your comments are welcomed. Kind Regards Zoltan
by on ‎02-21-2012 07:44 PM

From an engine/generator standpoint, I like to see a 100% load bank test (resisitve/reactive) at least once a year, preferably at the highest ambient conditions.  This allows a a very good test of the engine/generator and it's support equipment, such as cooling radiators, fuel supply system, etc.

 

But the engine/generator is only a part of your overall emergency power system, the ATS, and emergency circuits should also be regularly tested, but since most connected loads are usually lower than the rating of the engine/generator, this testing doesn't assure the unit is tested for maximum capability.

 

The reality for loads like data centers is that computer floor managers are usually not very happy to swap the power over to generator on a regular (or even irregular) basis, so you have to do testing in a way that usually doesn't affect power to the racks and support equipment, of course it doesn't really assure your overall emergency power system is going to do what you ask of it.  Ideally the thought of regular maintenance and testing are actually included in the design phase of the project, and that adequate bypass and multiple feed arrangements exist to be able to maintain and test equipment as needed, unfortunately these are usually a cut to reduce cost, and pass along the headaches to those who will operate the system later on.

 

As for the pros and cons or permanent vs portable load banks, I have worked with both, so the answer is a qualified "it depends"

 

Permanent load banks,

Pros

Already have assigned space with consideration for exhaust hot air, don't have to connect temporary cables, can be operated anytime a need to test is required, long term offer lower maintenance cost over renting a portable (several years)

Cons

Consumes space and needs adequate ventilation, requires regular maintenance, are expensive and get utilized infrequently (hard to convince the money folks in my experience)

Portable load banks,

Pros

Usually belongs to someone else, they have to assure proper function and operation, no large captial outlay for purchase and install, frees up site space for more important equipment.

Cons

Needs a place to park near the load, temporary cables are connected-risk of damage, improper connections, possible risk to workers depending on access, sometimes they show up and don't work, not your problem except the time allowed for the test which may have been hard to get, can't always get one when you need it (depends on where you are and size).

 

Hope that helps, Mike L.

Announcements
Help us grow the Caterpillar Community: Invite a Friend

Meet Our Bloggers