Circulator - flow variation with Termovar

I learned something this morning which I have not seen in any post and which, as the data shows, results in an area of automatic variable flow circulator operation in conjunction with the Termovar return temp protection valve. This may be especially interesting for those who operate their systems with a system return to the boiler below about 155F. It likely will not be important for those who operate their systems with return temps above 155F.

I changed my system over from a plate hx system to direct system operation. Reasons would be another discussion. I let the system cool down to depressurize for draining the boiler, which had antifreeze, and re-filling with water. Tank is horizontal 1000 gal LP. Piping is 1" steel between the boiler and tank. For this test the boiler was feeding only the storage, no other heat draw. I installed a flowmeter on the line to determine gpm.

The first interesting thing is the effect of the Termovar on boiler output to system vs feedback to boiler to maintain return temp. The data clearly shows that below about 150-155F system return to boiler, the Termovar is diverting substantial boiler output in the feedback loop back to the boiler, and as system return rises above this, the Termovar is diverting little if any boiler output in the feedback loop to the boiler. Initial boiler output to system was only 3 gpm at 96F system return to boiler, and boiler output to system maxed out at 8 gpm as system return to boiler reached about 155F.

The second interesting thing, although I did not record this for each data point, is actual Tarm output to system as related to stack probe temp. The burns during this data collection ranged between 400-500F stack temp (near the end, not included in data points, stack temps rose to 600F before coming back down). Stack temps closer to 500 produced the higher btuh outputs, and stack temps closer to 400 produced the lower btuh outputs. The Tarm is rated at 140,000 btuh, and the data shows that this can be reached only with stack temps above 500F.

A third interesting thing is effect of change in speed of the Grundfos UPS 15-58 on gpm's. My install is a bit unusual as I have the 15-58 on the boiler return, as normal, and I also have a Taco 007 on the tank return to the boiler, two circs operating to supplement each other. With the 007 shut down, but flow through the 007, 15-58 output was: Lo - 3.3 gpm, Med - 4.75 gpm, Hi - 5.50 gpm. With the 007 turned on, output was: Lo - 7.0 gpm, Med - 7.6 gpm, Hi - 8.0 gpm. All data shown below was with both circs on and the 15-58 on Hi.

A fourth interesting thing is that at gpm's below about 7, very little mixing was occurring in the horizontal tank and a high degree of stratification was occurring throughout the tank. As gpm's rose above 7, it was evident that mixing was increasing. I measured tank temp at both ends, top and bottom. For others with horizontal tanks, it may be important to keep the flow gpm's below 7 to achieve maximum stratification benefits. I have a diptube near the bottom of the tank for return to system, and supply to tank is about 1 foot from the diptube with a horizontal injection into the tank along the top and away from the diptube.

The last interesting thing is that during this entire test, the boiler did not idle at all. I continued the test past the data shown. My last data point collected is the final entry below. The absence of any idling maximizes boiler performance.

I fired the Tarm and let it come up to about 160F temp before turning on the circs.

Initial return from tank: 96F Boiler output: 160
gpm to tank: 3.0 Boiler return temp: 130 Output: 96,000 btuh

Return from tank: 130F Boiler output: 164
gpm tank: 6.25 Boiler return temp: 140 Output: 106,250 btuh

Return from tank: 134F Boiler output: 165
gpm tank: 6.75 Boiler return temp: 145 Output: 104,625 btuh

Return from tank: 137F Boiler output: 167
gpm tank: 7.50 Boiler return temp: 147 Output: 112,500 btuh

Return from tank: 142F Boiler output: 168
gpm tank: 7.75 Boiler return temp: 150 Output: 100,750 btuh

Return from tank: 148F Boiler output: 172
gpm tank: 8.00 Boiler return temp: 152 Output: 96,000 btuh

Return from tank: 153F Boiler output: 176
gpm tank: 8.00 Boiler return temp: 157 Output: 92,000 btuh

Return from tank: 156F Boiler output: 183
gpm tank: 8.00 Boiler return temp: 160 Output: 108,000 btuh

Return from tank: 159F Boiler output: 181
gpm tank: 8.00 Boiler return temp: 160 Output: 88,000 btuh

Return from tank: 162F Boiler output: 183
gpm tank: 8.00 Boiler return temp: 1662 Output: 84,000 btuh

Return from tank: 165F Boiler output: 190
gpm tank: 8.00 Boiler return temp: 165 Output: 100,000 btuh

Return from tank: 168F Boiler output: 193
gpm tank: 8.00 Boiler return temp: 168 Output: 100,000 btuh

[break in data]

Final entry, as burn was ending and stack temp dropped to 350F:

Return from tank: 180F Boiler output: 190
gpm tank: 8.00 Boiler return temp: 180 Output: 40,000 btuh
Final tank temp: Top - 187F Bottom - 185F

There is a substantial jump in gpm output to system at the 140 point. Different Termovars have different set points, so a person can select a Termovar to provide the level of return protection desired. Also I suspect that there is some variation in the thermostatic operation. I didn't collect too much data at the early stage of the burn (from 96F return to 130 return. I suspect that the 140 point holds for a fair while at low return temps to provide a fairly high output to system.

Regarding varying circ speed, I was a little surprised that no idling occurred during this entire cycle. In this situation I'm not sure what benefit there might be to varying circ speed, especially because entire output was going to storage.

Which temperature is your Termovar rated at?

Click to expand...

I t is a 4440A-3 which is rated at 72C (160F), and I think the reason for the less than 160 return on an initial cold start with cold return is that I have the balancing valve set for very little feedback flow through the Termovar. This works really well when the boiler is at operating temp, but results in return a little low on cold starts with cold return. I thought before about increasing this, as I know the return temp goes up when I open the balancing valve bit by bit.

Do I remember correctly that you burn primarily low MC softwoods? Was this all one firebox load? ... Any idea how many pounds of wood altogether?

Click to expand...

Correct, low MC softwood (all pine). I added wood from time to time to keep the firebox burning close to full burn the whole time as I wanted to determine whether or not the boiler would go into idle. I did not weigh the wood.

How much time elapsed between starting the circulators and boiler return reaching 140F?

Click to expand...

I should have, but did not, log the time.

You plan on trying this test again with the Taco turned off to see if the 15-58 alone can keep up with the boiler output without idling? And at which speeds?

Click to expand...

I am going to run the test again, and I will try it with the Taco off, Grundfos probably on HI, log time, weigh wood, and include tank temps with every log event. Probably will log every 15 minutes.

I measure tank temp by 2 thermocouples at each end of the 19' tank, one on top and one on the bottom, fixed to the exterior metal and well insulated. Hot water injection is into the top of the tank about at the 1/3 point from one end with a horizontal fitting to direct the hot water towards the 2/3 portion of the tank, with cold water output from a diptube just off the bottom and about 1' from the cold water injection.

I measure tank return temp and boiler output temp by two sensors fixed to the exterior of the respective pipes, about 8' from the boiler, with insulation over the sensors on the pipes.

I measure boiler return temp from a probe thermometer on the exterior of the boiler return line after the Termovar, and insulated again.

I'm thinking of logging the following data (feel free to make suggestions):

1) data measurements every 15 minutes
2) weight of wood at each load point
3) tank return temp
4) boiler output temp
5) boiler return temp
6) gpm
7) #1 end of tank: tank temp top and bottom
8) #2 end of tank: tank temp top and bottom
9) stack probe temp

Before running this test I may open a little the Termovar balancing valve to better approach 160 return temp from the start. I also will brush the hx tubes and clean the firebox of ash to maximize boiler performance.

PITA is my second favorite thing. The first is ... (from "One Night at McCool's")

I implemented a variation on Nofo's advice. The 3-speed on the boiler I wired into my differential controller, differential being between boiler supply temp and tank return temp. I then installed an aquastat on the boiler supply to control the Taco on the tank. I'm thinking that setting this for "on" at 180 would be about right. Now I can try the different speeds on the boiler circ for rapid boiler heating to 185 with low flow to system, and then at 185 the tank circ kicks on to increase flow to handle the high temp boiler output with maximum delta-T from the system. I'll pick the boiler circ speed which best "just handles" boiler output to avoid idling, thus keeping the boiler at high temp as much as possible.

Maybe a poor man's approach to an ECM circ setup and controller.

So what kind of flowmeter are you using? Is it 1”? The bigger ones that can take boiling water are pretty pricey from what I have seen so far.

Click to expand...

I's a Blue White 1" variable area rotameter. I'm pushing the max temp rating for what I have, but figuring I have a little bit of wiggle room. It is mounted on the return line from the tank so it is getting the lowest possible temp water from my system.

And how is it piped in? Permanently?

Click to expand...

I have it installed on a temporary basis. It has up to a 2 PSI pressure drop, so while my measurements are accurate when it on line, when I take it off I have to do a calculation for change in pump head and increase in flow. I think I can get pretty close to what it will be without the flowmeter.

The output on the Solo 40 is 1 1/4” pipe. Did you drop down to 1” to the tank for economy? Or some other reason?

Click to expand...

Yes - No. I initially had 3/4" line not knowing the high head and limited flow capacity of 3/4" pipe - ignorance is not bliss. I changed over to 1" due to economy plus I determined that I likely could get enough flow flow through 1" to handle the Tarm output with little or no idling. I think I'm going to be right on that. All my piping is indoors and the cost of changeover was not out of sight.

If you don't ask, you can't know. Or to be philosophical, to know and not do is not to know.

Here are the results of a more complete test run on a cold start. During this burn I kept the firebox about 1/3 - 2/3 full, adding dry pine as necessary, to maintain a pretty consistent high burn. I did not weigh the wood, which I will do on a subsequent test to measure a full load burn, beginning to end. I have a flowmeter installed on the tank return line. Comments:

1) This was a cold start, and I starting logging when I started the boiler with kindling.

2) The Termovar return protection valve clearly does its job, and it "modulates" circulation to system based on boiler supply temp and return water temp.

3) I have the Termovar balancing valve set for extremely low feedback flow, as is evidenced by the 140+ boiler return temp, although the Termovar is a 160F valve. On the next test run I will increase the feedback flow to achieve something closer to 160 boiler return temp minimum. The data then collected will show the impact of this change.

4) It took about 3:30 hours to achieve the first storage tank "turnover," that is, it wasn't until 4:30 that storage return temp (draw from bottom of tank) rose above starting (12:56) top of tank temp. That matches closely to an average gpm of about 4.25 during this period and assuming a little mixing in the tank. Stratification remained excellent at this flow rate.

5) The aquastat to turn on the tank circ was set for 180, and it actually went "ON" at a measured 176. I think I only want the circulation boost from the tank circ to come ON at a high boiler supply temp, so I likely will increase this setting for the next test.

6) The Boiler Supply/Storage Return temps are measured with the digital differential controller sensors; the Boiler Return temp is measured with a probe analog thermometer on the outside of the pipe and well insulated; the Tank temps are measured with a dual input digital thermometer with Type K thermocouples. All readings are +/- as good as the devices and sensors.

7) The GPM reading is to "closest 1/4 gpm" that I could determine. A float rises in the flowmeter tube based on the flow rate, and the tube is calibrated in even gpm's. The last reading of 8.1 is only because the float was a little above the 8 gpm mark.

Another test rust. Only change was to open the Termovar balancing valve a little more to achieve more quickly boiler return temp of 160F. As shown in the graphic, tank ended up at 190F. I burned 75 lbs of dry (est 7% MC) Jack pine. No idling occurred during the entire burn.

1) At 11:30 the controller turned "ON" the 15-58.
2) At 12:15 the aquastat turned "ON" the 007
3) At 3:15 the controller turned "OFF" the 15-58 as < 10F differential
4) Load burned out by 4:00

DaveBP said:

Jim, I assume your supply line into the tank points to the right end of the tank? That would explain how the bottom temps on that end are a little higher than the left end. But I'm impressed how even the top temps are left-to-right.

Click to expand...

That's right. I too was surprised at how even the temps developed and the degree of stratification that was maintained.

What shuts off the 007 tank circulator? When?

Click to expand...

That I have yet to install. The current plan is for an open on rise aquastat on the tank return line that will shut off the 007 at return temp of about 185F or a little higher. I have to think through what to do when the 007 goes off and the 15-58 also goes off because the differential closes, while a burn continues. I'm thinking of lowering the setpoint for the overheat aquastat so the excess heat will be dumped into the overheat loop before the boiler itself shuts down due to overheat. Any ideas?

Little lower BTUH rates this run. Whyzat?

Click to expand...

Two reasons, I think. Since I opened the Termovar balancing valve a little, flow rates were a little lower than before. Second, stack temps averaged a little lower. I'm going to do a couple of more tests with different conditions, and assuming I end up with some final settings a little different from what has been shown so far, I will put in another data table with the settings. In particular, after return temp rises above 160F, I want to close the Termovar balancing valve completely so that I can determine how much residual flow passes through the Termovar, as I assume there is some.

Anything else you would like to see?

I think you’re right that as soon as the tank return is much above the 140F minimum it would be good to shut off the balancing valve completely.
But how do you remember to open it again next burn?

Click to expand...

My preliminary thought on this would be to have an Automag type valve on the Termovar feedback line controlled by an aquastat close on rise or similar controller installed on the boiler return line before the Termovar. This could be set at 160, for example, so that at system return temp less than 160, no power to Automag and Automag would be open, allowing Termovar feedback. When system return reached 160, the Automag would be powered and close, terminating all flow through the Termovar. As system return again dropped below 160 less differential, power to Automag would be cut, valve opens, and feedback continues, as well as valve being open on next boiler burn.

Some months ago mid-winter I did shut off the Termovar balancing valve when system return was up to 160, and I did forget to reopen it the next time I fired the boiler, and I left for the burn to run on its own, as normal. Not nice. The fire took off, overheated the boiler, draft fan of course shut down, but a near full load of wood smoldered in the firebox for quite some time at high temp. A lot of condensation, black liquid draining down the chimney, and a mess on the floor. The creosote in the firebox burned off OK the next burn or two. I did not try that again.

Probably my last test run, unless something really new comes up or someone has a great idea that merits testing.

This test involved another adjustment of the Termovar balancing valve. I have a 1-1/4" gate valve as my balancing valve. On the March 09 test the valve was open one full turn; the March 10 test 1-1/2 turns; and this test is 1-1/4 turns. I also tried to maintain a pretty consistent mid-500 stack temp.

Observations:

1) The 3 gpm entries of 6.25 are not actual but are an estimate. Actual gpm's were alternating between 4.5 and 7.75 as the 007 turned ON and OFF based on boiler supply temp (actual ON is about 180 and differential OFF is 172). Thus, when it turned ON the inrush of cooler water brought down the boiler supply temp and cycled the 007 several times before boiler output maintained above 172 after a 180 ON.

2) Boiler idled first at 12:32 and back on-line at 12:48. It idled again at 12:55 and the idle continued through the last data point. The wood was nearly burned out at this point.

3) The drop in gpm from 7.75 to 6 is caused by the 15-58 going OFF based on BS/SR differential closing.

4) When both circs were ON and flow was 7.75 gpm with the balancing valve at 1-1/4 turns open, and boiler return (BR) at about 170, I opened the balancing valve all the way and flow dropped to 7 gpm. I then closed it all the way and flow increased to 9.5 gpm. This evidences that the balancing valve should be closed as much as possible to maintain the boiler return temp that you desire (probably 150-160 minimum) to maximize system flow. It also is evident that the Termovar never completely closes boiler feedback.

5) Based on 4), and with my system, I'm not sure whether a means to automatically further shut down the balancing valve and thereby increase flow and btuh would be beneficial, as my boiler goes into idle only at near maximum storage tank temperature. My system is operating quite efficiently with the Termovar functioning normally. This is particularly true because my flow will increase when I remove the flowmeter, which has a 2 psi max pressure drop from its operation.

I remember you mentioning fan-speed controllers for draft fans. Do you still use one to reduce your burn rate?
Does it work? Can you still maintain good gassification but at a slower rate?
Does it reduce the flue temps?
Do you get a longer burn time using your dry pine when you do reduce the burn rate?

Click to expand...

"Yes" to all questions. I use it just about only in cold winter weather (below about 10F) when I have an issue with excessive draft. As temps warm, I need full draft fan output to maintain good gasification. At times I reduce the speed of the draft fan really a lot and still maintain good gasification in the 450-550F stack temp range, which is where my boiler with my wood seems to work best.

With 1000 gallons pressurized storage is there any advantage to longer burn times since you capture all the heat to use later? I mean does loading fit the rhythm of your daily routine better.

Click to expand...

I did the 3 tests under pretty optimum conditions to maintain fairly consistent stack temps and get data which should be reasonably repeatable. This meant I didn't stuff the firebox and I added additional splits from time to time, and shook down the burning splits to maintain pretty consistent gasification.

Under real world conditions, there would be a fair amount of variability over the load-burn cycle, from start, to high burn, then burn leveling down, then coals to finish. In the real world I don't pay too much attention to anything but the end temp of the tank and how much stored heat I needed based on outside temp, or whether I'm going to be gone for a few days, and I will run a full or partial second or third load if needed. My approach is fire the boiler, leave and forget it. Pine does not burn a long time, probably about 4 hours is close to maximum, with most of the load pretty well finished by 3 to 3-1/2 hours. Also practically no coals left to start another fire sometime later. So I just load and burn as needed, as is convenient for me to do so, and as required by outside temps.

...So just to get academic, the only pattern I see corresponding to the stratification pattern is the circulator flow rate....

Click to expand...

I see your point, but perhaps there is a little different pattern than you suggest. First consider that this is a horizontal 1000 gal tank, about 39" diameter -- not much room for major stratification. But notice in each burn that, although storage return gradually rises during the early period, as gpm x hours starts to approach 1000, that is, when nearly a full load of hot water has been charged, the temp rises quite rapidly. Also notice how tank bottom temp, especially on the left end (opposite from the input direction of hot water), stays considerably under top temp except towards the end of the charge, when the whole tank is almost the same temp (Mar 10 and 15 tables).

The Mar 09 table probably would not represent a "normal" real world charge situation, as the tank start temp is really low (120-108) -- not much heat left here. And I did not continue this burn until the tank was fully charged.

The Mar 10 and 15 tables probably are closer to most real world operations, with tank start temps of about 160-142 (Mar 10) and 142-125 (Mar 15). This probably would be about the time that a person would consider charging the tank.

Glad you found this interesting.

I'm thinking about doing one or two more data collections, as a recent burn suggests that higher btuh can be achieved with the 15-58 on a speed lower than HI, with the 007 coming on at 180F boiler output. I may try a MED and then a LO setting. Trouble right now is that weather is warming, actually getting above freezing for a change.