Inverters for energy savings

I apologise for what will end up being a waffly post. 

I'm using my non-work account, but I have been a member here for many years.

Right. I have recently replaced some cooling towers and while I was at it, I had the control gear replaced. It used to be star-delta 17kW x 3 and I wanted to be able to adjust the speed so I specced an inverter per fan set, with a 0-10v ambient air temperature control, plus a load of other improvements to the plant, and I'm really pleased with it. 

I adjusted the parameters of the speed so it runs at 20% at anything below 10c and when it gets to 30c, it runs at 95% speed. 

Great, it works well, and I am getting the same cooling performance for our process. 

I then came across the first problem with it. I monitored the energy usage of the old towers, which were also 17kW x 3, and compared it to the new towers while they were running at about 50% speed, they are using one fifth of the energy of the old ones, they dont run for longer or anything, they just tick over for hardly any power. The problem with this is that I'm sure nobody will believe me, so I have kept this energy saving project to myself.

Anyway, I remain undeterred, and I have identified some plant that accounts for a significant load, two 22kW pumps, again, star-delta. now, when they were installed some years ago, one of them was pulling too much current, and we were told to throttle it back on the pressure side, and it is, it is only open 2/9ths of the way. The other pump has it's valves almost fully open. The delivery from them looks about the same - there is some restriction designed into the outlets to increase velocity. 

The easy thing for me to do is to have a panel built, inverters etc but there isnt really a reference to control the speed, so it would be set up at a fixed speed that still gave sufficient flow and velocity, maybe then having the valve open fully. 

I think I'm questioning it a bit, would it just be better to get two new pumps sized correctly? They arent cheap, one pump would be more than the inverters. I still cant quite get my head around why closing the pressure side valve reduced the current draw on that pump. Even if we only mananged to turn the pump speed down by 5% would be a decent saving. 

Finally, the pumps arent ancient, but they arent brand new, has anyone had experience of doing anything like this and what was the outcome?

  • Hi, Sounds like the first set up is working well.  Why not allow their doubters to have their day and open your results up to their scrutiny?  In the worst case they'll flag up something that you have missed and a lesson will have been jointly learnt.  

    In terms of your results, it sounds like you have some old and new systems operating together.  Are they on the same system?  If so, can you be sure the duty is being shared equally?  If they are seperate systems, how do you know the heat rejection demand from both sets of towers is equal?  

    In relation to the scond point - closing a valve at the discharge of a pump will increase the pressure at the pump discharge unless it is inverter controlled to deliver a set pressure and the transducer is across the pump.  In this case the pump will throttle back to deliver a lower flow and thereby lower it;s pressure to the set-point.   

    The question for me in this situation is what the pump is serving?  Is it a process reuqiring a set flow rate?  in whiuch case, what is the low -flow threshold before the process at the end of it starts to think about shutting down?  Or is it a variable flow, e.g. due to varying demand, in which case, what are the high and low limits 

  • Hi,

    Thanks for the detailed response.

    Last time I presented figures like this, I was laughed at for using pi, when calculating the area of a circle, it was in a report, so all the figures were removed - this part is a bit of a different story, I just take personal satisfaction out of these side projects.

    We do have a series of separately controlled equipment. The duty is shared because of how the pipework is configured, they are effectively a battery. There are quite a few variables, tonnes of material, surface area of material, temperature of material, ambient temperature, ambient humidity, start temperature of coolant. What I do is compare similar tonnage, material temperature and ambient temperature vs energy consumed. I have also just monitored total energy usage over several weeks vs the same weeks in the previous year, it always comes back at about 80% saving. I even went and put my clamp meter on the kit for a sanity check and that too confirms the difference. 

    Regarding the pumps, we dont have transducers on at the moment. They serve a process, so no long pipe runs, just several meters. I have got to grips with the reason the valve is closed to control the high current draw problem it had. The pumps basically squirt water out of a matrix of nozzles, so its always the same duty, nothing varies. I have established that 2/3 of the time, only a gentle operation is required so an inverter should easily be able to sort that via two user buttons. 

    The squirting is more of a subjective thing, is it sufficient rather than operating at a certain pressure or flow rate. There is a point at which it affects product quality, so that would have to be a bit of trial and error.

    I think what I would like to do is install inverters, open the throttled valve, modify the nozzles to improve the squirting performance, We did look at (I forget the proper name) venturi nozzles that entrain the fluid around the submerged nozzle, amplifying the effect. Ideally, I'd like to sort the pipework but I don't have the time or budget for that, so its making the most of what we have.

  • I can believe an 80% saving - having what is in effect a big fan on all the time rather than only when needed is remarkably wasteful as it has to be sized for the hottest summer day and greatest heat load, two states that seldom coincide.

    On a much smaller scale we do the same with the electronics and have speed controlled fans. Not only does it save energy, the fan bearings last longer and filters block more slowly, so the maintenance interval is greater.... and yes we find on cold days the fans are off, so a 100% power saving  on those days !

    To your pumps, if the throttle valve is re-opened but the pump slowed by inverter drive instead, you should be able to reach the same nozzle flow point, but without wasting energy in turbulent flow at the restriction, so the fluid, pump and pipe will all be ever so slightly cooler. 

    Ideally you'd have something measuring the speed of the fluid in the pipes and closed loop feedback but the current throttle valve is open loop, and seems to work, so maybe the inverter can be the same if its not that critical and just needs to be 'enough'.

    Again, from the point of view of service life having a slightly oversized machine loafing along running cool, rather than a small one struggling near to failure, is preferred, but only within reason of course - a factor of 2 perhaps..

    Mike.

  • I hadn't even really thought about the effect on service life. I was hoping to avoid the damaging effects of them being star-delta, such as them restarting when the water flow is reversed, I did that to some extent with non-return valves years ago, and, of course all the other bits that new inverters do like ramping the speed up etc. I noticed in the new inverters we fitted, you can set the natural frequency of the fan so the inverter will avoid it so it won't rattle itsself to bits.

    Measuring flow in pipes isn't that easy, and they are easily the cost of the inverter, I think that open loop is probably the right choice with it, as once it's set up, it can be left well alone.