Storage circulator speed?

[ohbie1]

I just got my Tarm Solo Innova 30 boiler up and operating on 2/28. ( Just in time for some single digit weather) I am using it with a 1K gal. wooden box storage tank. The unit comes with a Termovar circulator that has 3 speeds. Of course there is no info. on best speed or how to determine appropriate speed to use. I'm guessing that when storage is far below boiler temp., maybe top speed is most efficient, and as storage gets closer to boiler temp, slowing the pump down might make for better transfer. Does this sound logical?......or am I on the wrong track completely?

 

[BrownianHeatingTech]

You can measure the energy transfer by timing how the temperature changes in the tank, in a given amount of time (say, one hour).

Start on low speed, and see how it performs. The big thing to watch out for is that the boiler does not hit high limit and go idle. If it hits the high limit, you know for a fact that your flow rate is too low, so you should change to a higher setting.

Record the temperature when you start and, assuming that it does not hit the high limit, record the temperature an hour later. Next day, try to fire when the tank is at a similar temperature, and do the same thing, at the next speed.

Then compare which setting got you the best temperature rise in your tank.

Ideally, you will run at the lowest speed that you can without idling the boiler, but it's good to have records of whether a higher speed can get you better transfer, with your setup. That way, if you know that you're in a hurry to re-heat the tank, you can turn the circulator up for that particular burn.

Joe

 

[Nofossil]

The only thing I'll add to Joe's excellent post is that the colder the incoming water, the lower the circulator speed that you'll need. That's because it will require fewer GPM to carry away the heat that the boiler is producing.

 

[BrownianHeatingTech]

The only thing I'll add to Joe's excellent post is that the colder the incoming water, the lower the circulator speed that you'll need. That's because it will require fewer GPM to carry away the heat that the boiler is producing.

Yup, but that will require variable-speed pumping, to take advantage... Probably overkill, for most systems. Except when the system is owned by a boiler geek, I suppose!

Joe

 

[jebatty]

The flux capacitor does all of it automatically. The future is now.

 

[Nofossil]

The only thing I'll add to Joe's excellent post is that the colder the incoming water, the lower the circulator speed that you'll need. That's because it will require fewer GPM to carry away the heat that the boiler is producing.

Yup, but that will require variable-speed pumping, to take advantage... Probably overkill, for most systems. Except when the system is owned by a boiler geek, I suppose!

Joe

I'm setting up a test cell - I think I have a good industrial-grade solution to add variable speed control to my controller for standard Taco and Grundfos circulators for less than $100 per circulator. Does anyone know how slow you can run a circulator and still get reliable performance - 50% of rated flow? 25%? 5%?

 

[tuolumne]

Usually I am around when I am burning the boiler, so it is easy to adjust my Grundfos 1558 by hand to optimize charging efficiency. Nothing like the old human brain for decision making! OK, I'll admit that someday I'd like to automate all of this...for free of course.

 

[2.beans]

The only thing I'll add to Joe's excellent post is that the colder the incoming water, the lower the circulator speed that you'll need. That's because it will require fewer GPM to carry away the heat that the boiler is producing.

Yup, but that will require variable-speed pumping, to take advantage... Probably overkill, for most systems. Except when the system is owned by a boiler geek, I suppose!

Joe

I'm setting up a test cell - I think I have a good industrial-grade solution to add variable speed control to my controller for standard Taco and Grundfos circulators for less than $100 per circulator. Does anyone know how slow you can run a circulator and still get reliable performance - 50% of rated flow? 25%? 5%?

i dont know how low you can go but im in.

 

[jebatty]

The flow charts for the multi-speed circs should give you a good indication. A UPS 15-58 shows, for example, 2, 6 and 9 gpm at the same head of 10', but we know that head is roughly a squared function of flow, so this is misleading.

I suggest getting either a flowmeter, which also adds some head of its own, or using something like a Taco 1400-40, which has tappings on the inlet and outlet flanges, install some pressure gauges, and measure flow change by the change in pressure differential at different speeds.

 

[Nofossil]

The flow charts for the multi-speed circs should give you a good indication. A UPS 15-58 shows, for example, 2, 6 and 9 gpm at the same head of 10', but we know that head is roughly a squared function of flow, so this is misleading.

I suggest getting either a flowmeter, which also adds some head of its own, or using something like a Taco 1400-40, which has tappings on the inlet and outlet flanges, install some pressure gauges, and measure flow change by the change in pressure differential at different speeds.

I was contemplating the low tech approach - pump between two buckets with a 10' height difference, adjusted for the piping head loss. Instrumentation would be a stopwatch. Any reason why that wouldn't work?

 

[jebatty]

I think it would. What gets my brain fried is trying to translate performance based on one set of data to a different data set, the what if's for different situations. Maybe there's a computer program do this; I end up using Liquid Friction Loss on a trial and error basis along with the flow charts to make these translations.

 

[ohbie1]

Ideally, you will run at the lowest speed that you can without idling the boiler, but it's good to have records of whether a higher speed can get you better transfer, with your setup. That way, if you know that you're in a hurry to re-heat the tank, you can turn the circulator up for that particular burn.

Thanks for the easy test method. I do not understand why I would want to run at the lowest speed, and why I would not always be in a hurry to re-heat the tank. As long as I'm running the boiler, shouldn't I always want to get the most heat out of it the quickest way possible, otherwise isn't it wasting BTUs from the burn?

 

[DaveBP]

I do not understand why I would want to run at the lowest speed, and why I would not always be in a hurry to re-heat the tank. As long as I’m running the boiler, shouldn’t I always want to get the most heat out of it the quickest way possible, otherwise isn’t it wasting BTUs from the burn?

You want to get all the heat out of the boiler, yes. But once you have enough flow through the boiler to extract the heat that it is putting out any higher flow rate than that is not getting any more heat out. It's just using more electricity. Those extra watts can add up to hundreds of dollars over the life of the system. And that's wasting BTUs out of your wallet.

 

[WoodNotOil]

I experimented with my tank charging pump (Grundfos 15-58) through my plate hx on all three speeds at various temps. Not surprisingly I found the medium speed to be the good compromise setting...

 

[BrownianHeatingTech]

Thanks for the easy test method. I do not understand why I would want to run at the lowest speed, and why I would not always be in a hurry to re-heat the tank. As long as I'm running the boiler, shouldn't I always want to get the most heat out of it the quickest way possible, otherwise isn't it wasting BTUs from the burn?

Dave got it right. Running the pump too fast just burns up electricity.

As long as you are running fast enough to keep it from hitting the setpoint and idling, you are getting an efficient burn.

Joe

 

[in hot water]

Check out the solar differential controllers also. Many have 2, some 3, variable speed outputs. One controller could run 2 or 3 different circulators on a delta t or thermostat function. Boiler circ, system circ, storage circ, etc. I'm seeing them online in the 2-3 hundred dollar range. That is a good deal for that much control, most data log have PC connections and a lot of other features.

The Watts Radiant control with 3 outputs and a host of other features should be available this month. I've heard estimates of $250 to the contractor for a 3 output, variable speed, with data logging ability and btu monitoring. Resol is another brand to look at, it is private labeled under many of the solar brands like Oventrop, Caleffi, Schuco, Viessmann, etc.

For lowest power consumption, more and more ECM circs are hitting the market bring the price down. Laing and Wilo are out now, Grundfos and Taco should be very soon. Typically a 40% less power consumption with ECM based circs.

hr

 

[TimCroft]

My experience is that the hotter the boiler loop runs, the more turbulence you get in your 1K tank which mixes up the thermal stratification. The loop in the tank should run to the bottom of the tank first, heating the lower water hard and causing a thermal "swirl". A cooler flow from the boiler will gt you there but if it is about saving energy, then dumping that heat fast and efficiently is the game. I run Grundfos 15-58s with three speeds and the medium setting with my tarm 2000 excel is just about right to keep the boiler on the 'torque peak". A cool boiler jacket promotes creosote production which only decreases heat transfer over time.

 

[ohbie1]

You can measure the energy transfer by timing how the temperature changes in the tank, in a given amount of time (say, one hour).

Start on low speed, and see how it performs.

I started testing this morning. Even though I have not yet controlled all the variables, the initial test shows quite a difference between speed 2 & 3. (I have yet to test the slowest speed)

Degree rise in 1 hr.:

speed 2.......9.7
speed 3.......6.9

 

[jebatty]

I started testing this morning. Even though I have not yet controlled all the variables, the initial test shows quite a difference between speed 2 & 3. (I have yet to test the slowest speed)

Degree rise in 1 hr.:

speed 2.......9.7
speed 3.......6.9

I must be missing something here. Without considering heat losses, it takes 80,898 btu to raise 1000 gal of water 9.7 degrees (1000 x 8.34 x 9.7). At assumed delta-T boiler input/output of 20, that means at speed 2 you are moving 8 gpm (8.34 x 8 x 60 x 20). Similarly, it takes 57,546 btu to raise 1000 gal of water 6.9 degrees, which equates to 5.75 gpm at speed 3.

I take it that speed 3 is faster than speed 2, and if so why are you moving fewer gpm's at speed 3 than speed 2? Are there a number of other variables you are not considering or am I making a mistake?

 

[DaveBP]

At assumed delta-T boiler input/output of 20

The traditional delta-T=20F rule of thumb is hovering on the window ledge around here.

That rule makes for simple arithmetic calculations but also oversimplifies the physics to the point of misunderstanding. It implies that there is a linear relationship between water flow rate and heat delivery rate. If ohbie's return temp to the boiler is 140F it's entirely possible to get 190F coming out. The slower the water goes through, the more likely it will have time to pick up the BTUs. That's delta-T=50F. But if he plumbs up a monster circulator and increase the flow rate through the boiler by 1000% that delta-T will not be 50 anymore. And he won't be getting 10 times more BTUs.
We're on one of those algebraic teeter-totters here but the plank isn't straight. Higher GPM= lower delta-T. Lower GPM= higher delta-T. Where the best combination of delta-T and GPM is to produce the most BTUs/hour will vary with each system and most likely can only be found experimentally.

Get busy you data loggers. Don't be shy about posting your results. I, for one, just love being baffled.

 

[leaddog]

I must be missing something here but it seems to me if a boiler is making say 100,000 btu's and as long as it doesn't idle it does'nt matter how fast you pump the water you will tranfer those btu's. Now it would make a difference on the temp of water leaving the boiler. If you make 100,000btu's and not idle then 100,000btus must transfer so once you have the storage up to temp it will take the same time. I can see where you will mix the tank and the top won't be as hot but I think if you make 400,000btu's in 4hr's you will have 400,000btu's in the tank. Am I wrong here?????????????????
leaddog

 

[ohbie1]

I must be missing something here but it seems to me if a boiler is making say 100,000 btu's and as long as it doesn't idle it does'nt matter how fast you pump the water you will tranfer those btu's. Now it would make a difference on the temp of water leaving the boiler. If you make 100,000btu's and not idle then 100,000btus must transfer so once you have the storage up to temp it will take the same time. I can see where you will mix the tank and the top won't be as hot but I think if you make 400,000btu's in 4hr's you will have 400,000btu's in the tank. Am I wrong here?????????????????
leaddog

I felt the same way (that the higher speed would transfer more heat), until someone mentioned that their speed 2 was the most efficient. I have run two more tests since yesterday on speed 2 and had a rise of 9.7 and 9.0. So the speed 2 rise was not a fluke. I will run a few more on top speed shortly.

 

[DaveBP]

I must be missing something here but it seems to me if a boiler is making say 100,000 btu’s and as long as it doesn’t idle it does’nt matter how fast you pump the water you will tranfer those btu’s. Now it would make a difference on the temp of water leaving the boiler. If you make 100,000btu’s and not idle then 100,000btus must transfer so once you have the storage up to temp it will take the same time. I can see where you will mix the tank and the top won’t be as hot but I think if you make 400,000btu’s in 4hr’s you will have 400,000btu’s in the tank. Am I wrong here?????????????????
leaddog

Hard to argue with that logic.

Unless you had an outboard motor in your tank mixing it furiously so that it was the same temperature throughout, measuring its temperature at one point won't tell you how many BTUs are in there. In a stratified tank you need to measure the temperature at more than one point vertically to get an idea. The more points the more accurate your estimate will be; but it's still an estimate. And you'll need some differential calculus to do it right.

I think what ohbie is measuring is more like the delta-T at which he's transferrring the BTUs out of his tank rather than how many BTUs/hour he's getting. You can get them out of there at different temps. Slower pump, less electricity, more efficient. But the same number of BTUs, perhaps.

Here's a tank with enough sensors to get a reasonable estimate of the number of stored BTUs (or whatever they call BTUs in Sweden):

http://www.elstyrteknik.se/pannrum/

 

[BrownianHeatingTech]

I must be missing something here but it seems to me if a boiler is making say 100,000 btu's and as long as it doesn't idle it does'nt matter how fast you pump the water you will tranfer those btu's. Now it would make a difference on the temp of water leaving the boiler. If you make 100,000btu's and not idle then 100,000btus must transfer so once you have the storage up to temp it will take the same time. I can see where you will mix the tank and the top won't be as hot but I think if you make 400,000btu's in 4hr's you will have 400,000btu's in the tank. Am I wrong here?????????????????
leaddog

If you had a fixed temperature situation, that's correct.

However, as the temperature changes, the delta-T over the boiler changes, so the most effective way to transfer heat changes. The idea of setting the circulator speed is to achieve the most effective heat transfer, over the actual working temperatures of the system.

Joe

 

[ohbie1]

I think what ohbie is measuring is more like the delta-T at which he's transferrring the BTUs out of his tank rather than how many BTUs/hour he's getting. You can get them out of there at different temps. Slower pump, less electricity, more efficient. But the same number of BTUs, perhaps.

DaveBP is on the money. I did not mean to imply actual BTU transfer. I'm only using my top temp. sensor, and I have not done any Calculus for some 40 yrs. I am looking for highest transfer efficiency. My assumption is that whichever speed increases the temp. at that sensor the most, on a consistent basis, at various times of the day, and starting at a variety of temps. is my most efficient BTU transfer speed. Again this afternoon in back to back tests speed 2 beat speed 3 by 3 degrees per hr. (just at my top sensor)