Variable Speed Pump Data/Discussion

[brajgreg]

This could be long--sorry in advance.

I had a new pool installed last year, and I decided to go with a variable speed pump since it was for one reason or another free for me to upgrade to it. The pump itself is a Viron P280 variable speed pump with speed settings ranging from 1375 RPM to 3075 RPM. Last season, I noticed that occasionally the pump would blow sand (or what I think is sand) back into my pool. When I vacuum on filter, it spits it right back into the pool. I vacuumed on waste, and eventually, I had more in the bottom of the pool. I thought maybe I had cracked laterals, but that doesn't appear to be the case. Recently, I've been doing some research to figure out whether or not this pump, at 3075 RPM (which is it's automatic priming RPM when it turns on each time for 30 minutes), is giving too much flow for my filter. My filter is a 25" Cooper R series--I've attached the Data for it (the blue-ish background table. Any of the 25" models have the same specs, and if I'm reading it correctly, they have a max flow rate of 62.8 gpm. That's the first important number.

Now to the pump and figuring out my flow rates at different RPM's. I was unable to find Viron P280 pump curves online, but I did get in contact with them and *eventually* :grrrr: I was able to get some information. Their pump curve data shows 3 speeds (1375, 2000, and 3075 RPM) as well as power consumption and something else that I'm unfamiliar with. I would attach it, but the I have file size restrictions. I've basically mimicked their data in my Matlab chart anyway (see below). I wanted more speeds than they supplied, so I went into Matlab and did some math (fit a second order polynomial equation to their data, then plugged in a different Y intercept--this gives a decent approximation). I've plotted that data and attached it.

And last, but not least, I've hooked up a vacuum gauge and pressure gauge to my pumps suction and return sides, and I measured those as I went through several different RPM settings on the pump. Taking those two numbers, I multiplied the vacuum side (in HG) by 1.13 and the pressure side (psi) by 2.31 to get the TDH column. Then it's simply go to my Matlab chart, go up on the Y-axis to my RPM setting, along the curve until my head value and flow intersect. I've listed the flows in my excel data that I've attached as well.

So, now that I've probably got about 3 readers remaining , I come to the question portion.

(1) Am I right that the 62.8 gpm is my max flow rate for that size filter? It just says flow rate on the chart, not max. I haven't called the people at Cooper. I had been assuming it was max until I started writing this so I thought I'd ask.

(2) Is there anything you see wrong with my methodology? Am I correct in assuming that if my pump is on at 3075 RPM and thus flowing at 80 gpm, I may be overloading my filter and forcing sand into the pool?

(3) You may notice in my excel data that I'm getting 58 GPM at 2000 RPM, but dropping down a few RPM to 1925 and even to 1800 gives me 60 gpm. I guess that's due to better efficiency in that area possibly? The curves are leveling off down in that region. Also, as I'm typing this I just realized that those values are from my Matlab data, which likely has some amount error the farther we move to the right on the curves. Those RPM's and flow values are reasonably close, though.

(4) Also, at the low RPM settings my pressure side goes to 0, but I am getting flow from my returns in my pool. Do I need some minimum amount of pressure for my filter to run most efficiently?

Thanks for taking the time to read. I appreciate any responses or discussion.

 

[cowboycasey]

(1) Am I right that the 62.8 gpm is my max flow rate for that size filter? It just says flow rate on the chart, not max. I haven't called the people at Cooper. I had been assuming it was max until I started writing this so I thought I'd ask. I think your right on in thinking Max flow rate

(2) Is there anything you see wrong with my methodology? Am I correct in assuming that if my pump is on at 3075 RPM and thus flowing at 80 gpm, I may be overloading my filter and forcing sand into the pool? you may be right

(3) You may notice in my excel data that I'm getting 58 GPM at 2000 RPM, but dropping down a few RPM to 1925 and even to 1800 gives me 60 gpm. I guess that's due to better efficiency in that area possibly? The curves are leveling off down in that region. Also, as I'm typing this I just realized that those values are from my Matlab data, which likely has some amount error the farther we move to the right on the curves. Those RPM's and flow values are reasonably close, though.

(4) Also, at the low RPM settings my pressure side goes to 0, but I am getting flow from my returns in my pool. Do I need some minimum amount of pressure for my filter to run most efficiently? going to 0 is a good thing as long as there is still flow coming back into pool

Thanks for taking the time to read. I appreciate any responses or discussion.

I have responded in bold, I hope this helps

 

[pooldv]

You are on the right track there. My 3hp VS pump is too big for everything at max speed. I filter, skim and make chlorine at 1100 rpm (150 watts), I run at this speed 12 to 24 hours per day. Lower flow through your filter will result in more efficient filtration. I run at 1950 rpm (550 watts) for solar and/or waterfall. And I run at 2500 rpm when I want lots of flow through skimmers or bottom drain for brushing stuff to the drain. I never run at max speed, 3450 rpm, ever. The flow is just silly.

 

[brajgreg]

I've gone over the data some more, and added in the power calculations. I've attached what I have now. For reference (in case the picture eventually disappears or is corrupted as happens in forums sometimes), the pump at max speed--3075 rpm--costs $12.16/month for 1 turnover per day and flows at 80 gpm with a total dynamic head (TDH) of 27 ft.

At the minimum speed I've measured--1800 rpm which is a speed reduction of 41.5%--I get that it costs me $3.44 (a 71.7% decrease) per month to run the pump for 1 turnover per day and I'm getting a flow rate of 60 gpm (a 25% decrease) with a TDH of about 1.4 (a 94.85% decrease).

Places for possible error are:

1. My TDH calculations pressure side (and even the vac side but to a lesser degree)--getting in the lower pressure ranges on the gauge, it becomes increasingly difficult to know what I'm getting since I have a 0-30 psi gauge.
2. My pump curve data I received from the manufacturer only had 3 RPM ranges and 3 wattages for those ranges. I tried to be as accurate as I could using their data, but honestly their chart is just an absolute mess. I can't imagine an engineer making such a mess like that (note: I'm an engineer). Anyway, I used Matlab and then excel to do some curve fitting and create curves for the other RPM's listed in the chart. My power consumptions are all worst-case-scenario for each chart, which is not where I lie on the pump curve actually, but I thought it'd be a good place to put in some conservatism.

That's all for now. If anybody else has any input on my findings or questions, I'd love to chat about it. What this is showing me is that my pump has excellent flow even at lower speeds due to what appears to be inherently low TDH of my system. And while I was surprised at the flow rates, I'm really surprised at how little it'll cost me to run my pump at low speed vs the higher speeds. It isn't costing me much no matter what I decide, but the higher flow rates are too much for my filter anyway so I need to keep it running slower. The pump may never actually pay for itself in these circumstances, but I didn't have to pay for the upgrade anyway so that's good for me I guess!

I still have a little bit of measuring I want to do as a sanity check, and I'll probably come back once I've done that. I have some coiled up black flexible pvc on my concrete to act as a solar heater, and I can divert all flow to it. I'll measure the TDH and go to my curves and see what they say I should get. Then I'll fill up a bucket and see what I actually get over some period of time and compare. Hopefully they're within at least 10% of each other.

Cheers!

 

[mas985]

Just a couple of suggestions:

1) Measuring head loss at low RPM will result in very high error because the gauges do not have very high accuracy at low pressure/vacuum. So I found that you are much better off calibrating at high RPM and then just scaling the flow rate by the RPM. Flow rate is linear with RPM so if your measurements are not tracking that, then you have error in the measurements.

2) Both head curves and plumbing curves are very close to parabolic so you can usually fit those to the following equations:

Pump Head Loss = HM * ((RPM/3450)^2 - (GPM/QM)^2) ; HM is the head loss at 0 GPM at full speed and QM is the flow rate intercept at full speed.

Pump Wattage = Co + C1 * (RPM/3450)^3 + C2 * GPM * (RPM/3450)^2

Plumbing Head Loss = Cp * GPM^2 ; Cp is the plumbing curve constant.

These are the same equations that are used in my pump operating point spreadsheets (see sig).

 

[brajgreg]

Thanks for reading and responding, mas. I need to go read your signature posts. They look thorough and educational.

I don't think that RPM and flow rate are linear as the change in TDH and RPM are not linear and the pump curves are not linear. When you reduce your rpm, you reduce your head by some amount (which can be approximated very closely with a second order polynomial--R=.99 or better), and move down the curve some amount which gives quite the difference in flows. For instance, at 3075 RPM and a TDH (measured) of 27 ft, I get right at 80 gpm from the manufacturer's data. If we put in 3075/2 = 1537.5 RPM, I don't get half the flow rate at 40 gpm. If my head was as high as 3 ft (it's not at that flow rate, but let's say it was) or as low as 0 (it's also not 0, but this gives us a good bound for the problem), I get a flow rate between 48.75 and 54.8 GPM which is about 60-70% of the 3075 flow rate.

For reference, in order to get about 40 gpm, my TDH would need to be 6.75 ft which is well within the readable range on my setup of gauges. It's well below 6.75 ft.

And as for your second point, I may compare your numbers to mine over the weekend if I get a little time--with the assumption that in the first equation your 3450 would be my 3075 or Max RPM. My curves were generated using the supplied data from the manufacturer. I put their points in excel for the three speeds, created a trendline (R=1) using a second order polynomial, and used that polynomial equation (second order) to generate my curves. Similarly, I needed to find the y-intercepts, so I plotted their y-intercepts vs RPM on a chart, used a second order polynomial regression again, and using that equation I plugged in the various RPM's shown to find their y-intercepts. Using this technique I'm able to mimic their data within a few percent at almost every point so I feel my data should be accurate.

 

[mas985]

I don't think that RPM and flow rate are linear as the change in TDH and RPM are not linear and the pump curves are not linear. When you reduce your rpm, you reduce your head by some amount (which can be approximated very closely with a second order polynomial--R=.99 or better), and move down the curve some amount which gives quite the difference in flows. For instance, at 3075 RPM and a TDH (measured) of 27 ft, I get right at 80 gpm from the manufacturer's data. If we put in 3075/2 = 1537.5 RPM, I don't get half the flow rate at 40 gpm. If my head was as high as 3 ft (it's not at that flow rate, but let's say it was) or as low as 0 (it's also not 0, but this gives us a good bound for the problem), I get a flow rate between 48.75 and 54.8 GPM which is about 60-70% of the 3075 flow rate. .

Um, yes RPM and GPM are linear. I know this for a fact. It is dictated by the pump affinity laws:

Pump Affinity Laws

When you halve the flow rate, head loss goes down by a factor of 4. If you don't "get" 40 GPM, you are doing something wrong in your calculations.

For reference, in order to get about 40 gpm, my TDH would need to be 6.75 ft which is well within the readable range on my setup of gauges. It's well below 6.75 ft.

As I mentioned before, measuring at lower speeds is worthless. There is too much error. But 6.75' of head is exactly right. It is 1/4th of 27' which it should be. Your measurements are just wrong.

But if you still don't believe me, download the Energy Star pump measurement data and in the DB, they have examples of two speed and variable pumps which shows this linear relationship.

 

[brajgreg]

You're right. I think what I'm seeing is due to a likely issue with pressure side gauge. I've suspected it for a while anyway. This would be a good time to replace it. Thanks again for looking over my write-up.

 

[mas985]

A new gauge is unlikely to solve your problem. No standard pressure gauge that I know of has good accuracy at the low end. The accuracy is usually speced as a percentage of the maximum reading. So if you want higher accuracy at the low end, you need to get a gauge with a much lower max reading (e.g. 5 PSI max). Same for the vacuum gauge.

But it is worth repeating that will get better results if you calibrate at full speed and just scale for RPM.