Article 55: Pumps versus Donuts
Jim was happy. As a maintenance manager of a municipal water plant, he liked days like this. Sunny and pleasant. The plant ran smooth with nothing major happening, the pumps were pumping water where it was supposed to go, the day was coming to a close, and he was ready to take off home to take kind to a basketball game.
- Hey, Jim, - the guy on a phone sputtered, - did I catch you at a good time? Got a minute?
- Sure, Bob. But just a quick one. I am getting ready to head home. What’s up?
- Not much. Just wanted to know if you have some time tomorrow morning; I was gonna come by to talk about your pumps.
- No problem. Come on by. See you tomorrow.
The next morning, around 8am, Bob rolled in, with a box of donuts in hand:
- Jim, here is for the guys at the shop, to get the day started sweetly. How are the pumps doing?
- Doing fine, Bob. That spare mechanical seal you sold us is still in my office, ready to go if anything goes bad. But so far, knock on wood, all is well.
- Good to hear, Jim. I must say we make good pumps, do not fail too often. But I looked up my record and noticed we have not done any major repairs for you for rather long time now.
- And that’s a good thing, Bob. No problems is what we want! You guys do make good pumps.
- We sure do. But have you checked if your pumps are still efficient?
- They are very efficient. Pumping that darn water nonstop, like there no tomorrow!
- That’s not what I mean, Jim. I mean, do they burn too much energy?
- Burning energy? I have no idea. All I know they run well, no vibrations, no leaks, no trouble for me.
With that, Jim took another donuts:
- Just making sure I keep my weight on the level! – let’s get the rest of these to the guys at the shop, and then you can tell me all that fancy stuff about energy you are ready to clutter my brains with.
When Jim and Bob return back to Jim’s office, Bob pulled out the pump catalogue:
- See, Jim, these pumps are supposed to be almost 90% efficient, that’s according to our books. But its’ been a few years now, and they do wear and tear, and that makes them take more power than they should.
- Really? Just after four years? Do they wear out that soon? After all, it is basically clean water we are pumping.
- That’s true, Jim. But, you never know. Why not check it anyway? Listen, these are 3000 horsepower motors! – guess how much does it cost you to run these babies?
- Hey Bob, - motors-shmotors. What do I care? They run all day long, and do not fall over, and that’s all I care. And whatever we pay to run them, go talk to accounting, I am not a bean counter. I have enough problems with grass going wild by the front gate, and Charlie is after me for not watering it every week, - let him worry about this, what do you call it, efficiency thing or something?
Bob pulled his calculator:
- OK. But just for kicks: say your 3000 ho motor runs nonstop. That is (click click click…) about $2238 kilowatts, and in 24 hours 365 days, you get, let’s see… - 19,604,880 kilowatt-hours, which, at 10 cents per kilowatt-hours makes it nearly two million bucks!
- Wow, Bob. That’s a lot. You probably need to work a week and a half, a good week and a half, to make that kind of money! Good grief. So let them burn. Somebody’s paying. That’s what makes our economy going around! When will you, young folks, ever learn that!
Bob thought for a moment, and continued:
- True, Jim. You are right about that. I too have mortgage to pay. Any one of your pumps repair jobs will keep me going for a bunch of years! (just kidding!). But, listen, do you mind if we actually measure these pumps energy consumption? We signed up with this company that handles such things, they call it a “PREMS” technology, - Pumps Reliability and Energy Savings Monitoring System. It’s neat: we hook it up and you get your data live, nonstop, at any of your office computers!
After two donuts and a steamy cup of coffee, Jim was in a good mood today:
- Sure. Go ahead, do your measurements, set the system. Let me know what you find. If any good, I will tell Charlie, otherwise do not bother me any more today. I’ve got a meeting to go to. You can talk to my mechanic, Rusty, if you need any help with your equipment.
Bob went to action. Pulling his car to a main water booster pump, he and Rusty attached instrumentation to his device. Vibration and temperature transducers had mag bases and took no time to plop on the bearing housing, but for the pressure they teed off the existing gages. Flow and power already had output signals on the plant DCS controllers, and they connected output to that. The system was up and running just before lunch:
- Hey, Jim! - want to grab a bite to eat, and I will show you what I found? My system is collecting data from the instrumentations over there, and sends the gateway signal wirelessly to the cell. From there, we get our software process it and transform into a pump performance curve, and to compare it versus the original OEM curve. I did not have a chance to analyze it very much, but if you want we can take a quick look at it at lunch. I have my computer with me.
- Very well, mister pest, let’s go. Show me you magic.
While munching on a sandwich, Bob opened his laptop:
- See, we do not even need to bother your plant computer folks, it is wireless. We also have an option to connect to a local router, but going wireless is so much faster. Anyway, let’s look… The sold lines is the expected performance: pump head, power, and efficiency curves versus flow. The dash line is what we have already gathered this morning from your pump running. Rusty asked operators to throttle the valves in and out, so that the pump was forced to run at various flow rates, and the result is almost the entire flow range is covered (except we could not get too far out in flow, since your piping system limits it to about 40,000 gpm).
- Well, that was quick. You did all that in one day?
- Sure. Thanks to Rusty who helped with some wrenches, and the fact that you already have magnetic flowmeter, and we simply connected the 4-20 mA signal to its output DCS terminals. Now, let’s see. The BEP (best efficiency) is at about 35,000 gpm – and the original curve says about 40,000 gpm… You also lost some pressure (head is less from the original), and the power is up. Sounds like you might have some wear, maybe rings opened up, or some rub internally, which takes more power.
- Fancy, Bob. Anyway, how much all that costing us?
- Well, I added a tabulation at near the curve there. See, at 40,000 gpm you should have 255 feet of head, and you got 227 feet – that is 11% reduction in pressure. The power should be 2900 hp, and you are actually taking 3045 hp. Gee, seems like you are running in to a motor safety factor! I would not be surprised if you motor starts tripping pretty soon.
- What? Are you kidding? This is almost a brand new motor! – you are the one sold it to us about four years ago with the pump!
- We did, Jim. But, not sure if you know, we told your engineers at that time to upsize the motor a bit, - going to 3500 hp instead of 3000 hp, in case something like this happens!
- Asked engineers… - that’s your problem right there, anyway. What do these turkeys know about running a plant. I have been here 30 years, and yet to see any of these momma’s boys come get their hands dirty. Gee, don’t even get me started on this…
- Yeah, Bob, I know how you feel. I am an engineer myself, and sometimes feel I should be a baker. Anyway, all that aside, your efficiency at 40,000 gpm is 75.3% versus 88.8% it should be.
- Really? And so?
- And so, that is 13.5 % difference. Remember I told you, you are paying about $2M for this pump running non-stop? That means that, roughly speaking, it is $2M/100 = $20K per point of efficiency. For 13.5 points, that is over a quarter million dollars per year, running nonstop!
- But we do not run non-stop! We probably run about 10-12 hours per day on an average.
- Fine, than it is “only” $125K per year - wasted.
Bob took off his baseball hat, scratched his head, and replaced the hat backwards:
- You sure know how to turn a good day into a nightmare. Good thing at least I do not pay for these $125K. But Charlie sure would not be happy to hear that. But if I tell him, I know what he will ask.
- How much would it cost to fix the darn thing?
- Yeah, exactly. What would you need to do to fix it, and how much would it cost Charlie?
- Not that much. I can get you a quote tomorrow or Thursday.
- A ballpark?
- Well, based on our previous repair, that is probably new wear rings, probably a new shaft, bearings, unless the impeller also needs fixing, - which is a good idea while we are at it. I’d say, probably a $100K job.
- Hmmm… A $100K? And that would save us what did you say, a $125k a year? So, roughly a one year payback?
- That sounds right.
- Interesting. But you know, our budget is tight this year, Besides, I am retiring in October, and after that let a new guy worry. I tell you what, - let it rip. When it falls apart, we will call you. Thanks for coming, Jimmy. See you later. I got a meeting to go to.
1. Nelson, E., Maintenance and Troubleshooting of Single-Stage Centrifugal Pumps, Teas A&M Pump Symposium, 1984
2. Nelik, L., Let’s Find the True Best Efficiency Point (BEP), Pumps & Systems magazine, May 2015
3. PREMS-2A Pumps Reliability and Efficiency Monitoring System, rev. 2A, March 2015: