Plate HX

[jebatty]

I don't think there will be any icky water and sucking air issues with pressurized storage. Does pressurized storage also address the head issues with the 007?

 

[Perfect Heat]

I am in the process of building a heat storage tank in my basement. I would like to avoid running the piping through the liner if possible. Can one run the piping over the top and down inside the tank vice through the liner? I am planning on using a flat plate heat exchanger.

 

[WoodNotOil]

Jersey Bill - I am trying to understand and take into account some of your comments on plate hx. You are talking about 10 ft of head pressure drop from friction in going through the plate hx, right? Assuming that is correct, according to this chart: http://www.radiantheatproducts.com/store.asp?pid=5312&catid=10050 the taco 007 should have 20gpm at around 9 ft of head. You claim it to have 5gpm. Which is correct?

Also, I had not taken the rusting of pumps into consideration. On the open tank end are bronze pumps really required? Will the taco cast iron pumps rust and corrode? If so, then this becomes really cost prohibitted. For the Taco 007 the cast iron is around $58 and the bronze is $220. If I have to have two bronze pumps, a plate exchange, and two zone valves just on the open tank end, making copper coils might be more practical. Assuming this is neccessary, rough cost would be:

2- Bronze pumps - $440+
30 plate hx 225,000 btu - $200+
2 - Zone valves - $160+
Rough Total - $800+

Is this expense necessary?

 

[WoodNotOil]

message duplicated, sorry.

 

[jebatty]

I have to question corrosion of iron circ pumps on an open system, mainly because I operated a traditional OWB for 10 years, same pump, without a corrosion issue, and prior owner operated for unknown number of years before that. I did add boiler chemical periodically, maybe 3-4 time over the heat season. I can't say there was no corrosion, only that the pump continued to operate and no failures. Might be much cheaper to use an iron pump, even if you have to occasionally replace it.

 

[jebatty]

By way of concept, use plate hx, eliminate a zone pump, what about the following:
1) Boiler providing heat, no call from zone: P1 and P2 on, valves 2 and 4 open, valves 1 and 3 closed. All heat to tank, tank return to hx.
2) Boiler providing heat, call from zone: P1 and P2 on, valves 1 and 2 open, valves 3 and 4 closed. All heat to zone, zone return to tank, tank return to hx.
3) Boiler not providing heat, no call from zone: P1 and P2 off, valve 1 closed to prevent thermo-siphon, valves 2, 3 and 4 irrelevant.
4) Boiler not providing heat, call from zone: P1 off and P2 on, valves 1 and 3 open, valves 2 and 4 closed. heat from tank to zone, zone return to tank.

 

[Jersey Bill]

Sorry, I was out for a while...

No, with a closed system there wont be a problem with icky water or the suction side of the pump. thats only for an open system trying to skim the hot water off the top of the tank.

woodnotoil- i have the same pump curve, I dont see where you get the 20gpm@9'. If you follow the 007 curve, it crosses 20 gpm @2', 15gpm@6', [email protected]', 5gpm@9'.

Also, you would only need a bronze pump on the open side of the heatx. If the system is open, oxygen can get in. This will rust iron. when the system is closed, the water becomes anerobic (without oxygen) so iron cannot oxidize.

jebatty, you have an OWB on an open system and you say you dont have any corrosion problems with iron in your system? where is your system open? maybe chemical treatment is helping this issue.
also, in your last diagram, it looks like you are isolating the wood boiler with the heatx, not the thermal storage. But your control logic seems to work. I am concerned about the flow rates. In the case 1, all the heat from the boiler has to be pulled out by P2. with a 10 deg delta, thats 20 gpm, for 100 mbh. what size is the boiler? the pipe size would be 1.25-1.5". Then in case 4, the same pump, and 20 gpm is going to the zone. those are big and expensive zone valves. There is also a pressure drop through the valves, and in any case, 2 valves are in series. the heatx for 20 gpm is pretty large also.

I tried to find the pressure drop of the flatplate heat exchangers, but this data wasnt readily available. does anybody have the pressure drop curves for any heat exchagers?

 

[Nofossil]

I tried to find the pressure drop of the flatplate heat exchangers, but this data wasnt readily available. does anybody have the pressure drop curves for any heat exchagers?

I tried without success. I even tried plugging in various bogus values into an online HX selection tool to try and infer the head loss. From what I can tell, they have extremely small head losses. Tankless hot water heaters, however, present extremely high head losses - 40 feet or more at 6gpm in one case.

 

[WoodNotOil]

I spoke to a rep from www.radiantheatproducts.com today and got some clearer figures and answers on this type of setup.

Corrosion
He said there is corrosion in any system whether or not it is open or closed. Bronze pumps are only needed for potable water systems. Cast iron pumps are fine in both open and closed systems so long as it is not potable.

Flat Plate ft. of Head
In my design I want to transfer the 140,000 btu output of the tarm to the storage tank and be able to draw from the tank as well through one side of the plate. At 180* feed and 160* return temps, a 30 plate exchange is needed and it would have 12 1/4 ft of head. Because my storage tank water will be pulled up by a pump and through the exchange about 2-6 ft above the the tank, we brought the total ft of head to about 15'. He said I will need 15 gpm for the correct btu transfer on the 30 plate.

Pump Size
For 15' of head at 15 gpm you would need a Taco 0010. Since this is such an expensive pump, he suggested using the Grundfos 15-58 3 speed pump on its highest speed which is about $79. My application will need two in order to pump the tank side in either direction to maintain stratification with zone valves to isolate them. I happen to have two of those feeding my zones on speed 1 that I can replace with 007 pumps. (They were part of a radiant floor package I bought when retro fitting my 1st floor) Here is a rough estimate on what the tank side setup with plate hx would be:

30 flat plate hx - $200
2 - Grundfos 15-58 - $160
2 - Zone Valves - $160
Rough Total $520

I still think that beats the price of making a coil hx. What do you think?

 

[Nofossil]

Looks about right. Check eBay for prices.

If you're pumping out of and back into the same tank, there's no head involved from tank height. The lift to pull the water up is balanced when the water goes back down. You'll want to make sure that the pumps are below the waterline to avoid air getting trapped in the pumps.

How much heat transfer do you lose if you drop to 8 or 10 gpm? I don't think that plate heat exchangers lose much. Where you get into trouble is if you're trying for a very low delta T between the two paths.

 

[jebatty]

jebatty, you have an OWB on an open system and you say you dont have any corrosion problems with iron in your system? where is your system open? maybe chemical treatment is helping this issue.
also, in your last diagram, it looks like you are isolating the wood boiler with the heatx, not the thermal storage. But your control logic seems to work. I am concerned about the flow rates. In the case 1, all the heat from the boiler has to be pulled out by P2. with a 10 deg delta, thats 20 gpm, for 100 mbh. what size is the boiler? the pipe size would be 1.25-1.5”. Then in case 4, the same pump, and 20 gpm is going to the zone. those are big and expensive zone valves. There is also a pressure drop through the valves, and in any case, 2 valves are in series. the heatx for 20 gpm is pretty large also.

I think part of this was answered indicating iron pumps on an open system (non-potable) are OK - also consistent with my experience.

Regarding flow rates, etc., all I can convey is my experience. Tarm Solo Plus 40, 140,000 btu rating. Boiler side: Taco 009, 1-1/4 to 1" on hx, 5 x 12 x 30 plate. Tank side: Taco 007, 1" hx to 3/4" line to tank; 1/2" tank to 3/4" line to 1" hx from tank. The tank and boiler are 20' feet apart horizontally. Input to tank at about 44", output from tank at about 6" (bottom).

The following are approximates. Assuming tank input to hx is 80-130, boiler input to hx will hold around 170, hx output to tank with above inputs will be 140-155. Up to this point the boiler will be in high burn continuously with no idling. The hx system is stripping essentially entire boiler output.

As tank input to hx begins to exceed 130, boiler input to hx will rise slowly and hx output to tank also will rise. Generally there will be no idling until tank input to hx approaches about 140 and above. The boiler temp will gradually rise to 190 and cycling will begin. As tank inputs to hx rises to 150 and above, cycling will be regular. Also delta T between boiler and tank sides of the hx will close to between 5-10*. I never have achieved less than 5* delta T. I normally do not attempt to drive the tank to above 170 top of tank, and at 170 top, bottom will be about 160 or a little more.

It is apparent that the highest efficiency in boiler operation, as measured by lack of boiler cycling, is at hx input temps (tank/zone return) of less than 150. Cycling off times initially will be 8-12 minutes, and on times of 30-40 minutes, but the off time will increase and on time will decrease as tank input to hx rises.

All temp measurements except the tank top are with probe meat thermometers cable tied close to the source being measured and wrapped with insulation. The tank top is measured with a dial thermometer with a probe in a well in the top of the tank. The probe thermometers were calibrated to read the same temp as the dial temp at 140. In essence, temp measurements are consistent with each other, subject to error in each device, but all could be equally off if the dial thermometer is off, and each is subject to its own error.

 

[Donl]

I spoke to a rep from www.radiantheatproducts.com today and got some clearer figures and answers on this type of setup.

Corrosion
He said there is corrosion in any system whether or not it is open or closed. Bronze pumps are only needed for potable water systems. Cast iron pumps are fine in both open and closed systems so long as it is not potable.

Flat Plate ft. of Head
In my design I want to transfer the 140,000 btu output of the tarm to the storage tank and be able to draw from the tank as well through one side of the plate. At 180* feed and 160* return temps, a 30 plate exchange is needed and it would have 12 1/4 ft of head. Because my storage tank water will be pulled up by a pump and through the exchange about 2-6 ft above the the tank, we brought the total ft of head to about 15'. He said I will need 15 gpm for the correct btu transfer on the 30 plate.

Pump Size
For 15' of head at 15 gpm you would need a Taco 0010. Since this is such an expensive pump, he suggested using the Grundfos 15-58 3 speed pump on its highest speed which is about $79. My application will need two in order to pump the tank side in either direction to maintain stratification with zone valves to isolate them. I happen to have two of those feeding my zones on speed 1 that I can replace with 007 pumps. (They were part of a radiant floor package I bought when retro fitting my 1st floor) Here is a rough estimate on what the tank side setup with plate hx would be:

30 flat plate hx - $200
2 - Grundfos 15-58 - $160
2 - Zone Valves - $160
Rough Total $520

I still think that beats the price of making a coil hx. What do you think?

Would it be correct to assume that if a 30 plate HX has 12.5 ft of head that a larger, say 40 or 50 plate HX would
have less ft of head. Assuming that is correct even though a larger HX will not significantly increase heat transfer, it may allow a smaller pump such as a 007 to be employed. The savings in energy costs over time could offset the cost of the larger HX. Make any sense?

 

[jebatty]

Would it be correct to assume that if a 30 plate HX has 12.5 ft of head that a larger, say 40 or 50 plate HX would have less ft of head. Assuming that is correct even though a larger HX will not significantly increase heat transfer, it may allow a smaller pump such as a 007 to be employed. The savings in energy costs over time could offset the cost of the larger HX. Make any sense?

I think there is some confusion here. Or I'm confused. If you're operating a pressurized system, isn't it true there really is no head? It is a circulating pump, really not pushing or pulling anything, just causing a pressure differential which moves the fluid. If it's an open system, then there is head, right? Don't we need to distinguish between the two with regard to our comments?

It takes about 12 psi in a pressurized system to service a two story building. That amount of pressure will push water to about 25 feet, so if the system is 12 psi minimum in a two story structure, there is no head. See the following:
http://www.accontrols.com/documents/FeetHeadofWatertoPSI.pdf

 

[Nofossil]

Would it be correct to assume that if a 30 plate HX has 12.5 ft of head that a larger, say 40 or 50 plate HX would have less ft of head. Assuming that is correct even though a larger HX will not significantly increase heat transfer, it may allow a smaller pump such as a 007 to be employed. The savings in energy costs over time could offset the cost of the larger HX. Make any sense?

I think there is some confusion here. Or I'm confused. If you're operating a pressurized system, isn't it true there really is no head? It is a circulating pump, really not pushing or pulling anything, just causing a pressure differential which moves the fluid. If it's an open system, then there is head, right? Don't we need to distinguish between the two with regard to our comments?

It takes about 12 psi in a pressurized system to service a two story building. That amount of pressure will push water to about 25 feet, so if the system is 12 psi minimum in a two story structure, there is no head. See the following:
http://www.accontrols.com/documents/FeetHeadofWatertoPSI.pdf

If you have a pressurized system, there's no head as a result of the elevation. The effort required to push water to the second floor is exactly balanced by the water flowing back down from the second floor.

In order to prime the system in the first place, you would need a big enough pump to overcome the height. You will also have head loss from the flow restrictions of the various components in the loop. However, a basement zone and a second floor zone with the same components will have the same head loss.

 

[jebatty]

In order to prime the system in the first place, you would need a big enough pump to overcome the height.

Are you sure this is correct? As the system is pressurized by addition of water, water will fill the lower system first, forcing air to the top. With venting or a bleeder vent at the top, the compressed air will be released, allowing water to fill the entire system. The pump will not need to overcome any height, as pressure is equal throughout the system. At least this is what my "head" tells me.

 

[Nofossil]

In order to prime the system in the first place, you would need a big enough pump to overcome the height.

Are you sure this is correct? As the system is pressurized by addition of water, water will fill the lower system first, forcing air to the top. With venting or a bleeder vent at the top, the compressed air will be released, allowing water to fill the entire system. The pump will not need to overcome any height, as pressure is equal throughout the system. At least this is what my "head" tells me.

Exactly right. However, some folks don't have a properly located bleeder at the top. In that case you'd need a stronger pump, or you'd have to flush the air out of the system with house water pressure.

Even with a bleeder at the top, you'd have to ensure that your system was pressurized enough to force water to the top.

Worst, case, you'd need some help getting started. In any event, once air is flushed the extra height does not affect circulator selection.

 

[Donl]

I think there is some confusion here. Or I'm confused. If you're operating a pressurized system, isn't it true there really is no head?

If That were true then why would you need anything but the smallest pump in any pressurized system? From what I understand head is a result of gravity and friction loss within the system. In a pressurized system the gravity portion is negligible, however friction loss still must be taken into account. Size and length of pipe, number of fitting have to be considered. That is why in my question regarding head loss attributed to the HX I would think that a larger HX would have less head loss thus permitting a smaller pump. I guess I,m trying to understand if this idea is worth pursuing?

 

[Eric Johnson]

Just as an aside, you can easily raise the pressure in your system by manipulating the lever on the top of your water feed regulator, i.e., bypass the pressure regulator and allow water into the system at a higher pressure. I've found this approach useful for venting the radiators on the second floor of my house. One or two of them won't completely vent at normal system pressure (12-15 psi). When I do my annual rad venting at the beginning of the heating season, I bump the system pressure up to 20 psi to bleed the problem rads, then bring it back down into the normal range after the bleeding is done.

 

[Redox]

Head is head to a pump. It doesn't matter if it is to pump it uphill or around a loop or even down through a building. You will need so many PSID to move X amount of water. Only time you have to figure in height is in the fill pressure (2.31ft / 1PSI). Agreed, you don't need a large circulator to move a large volume as long as you pay attention to pipe sizes and HX pressure drops. Keep the lines and exchangers big and you save on pump horsepower. LEED certification is changing the way people think about pumping losses. NPSH (net positive suction head) is a different issue, but generally not a problem in residential applications. Plate heat exchangers have an extremely low pressure drop at most normal flow rates. I used to have some selection software from SWEP, but not sure where it is at the moment.

Eric, just a guess, but if you increased your fill pressure a little bit, you may not have to bleed at all. The air only gets there if it is drawn in by a vacuum or if you have a lot of fresh water exchange.

Even the ME's blow this on occasion. We have a new building here with 30 pound boilers and 4+ stories up that is pulling a vacuum on the rooftop AHU's. Waiting to see who ponies up the scratch for new boilers or heat exchangers.

Chris