There is a private function,
Click on the member link, I think that's how it works,
Or you e mail direct [email protected] I'll be there Tuesday, barring me winning the lotto...
Dan

If using a standard water heater for a buffer, cap the cold in and change the drain to a tee and put the cold in there. The element ports can be made into ports for the loops (or the boiler, if desired)

Posted By MikeSolar on 02 Sep 2012 10:20 PM
If using a standard water heater for a buffer, cap the cold in and change the drain to a tee and put the cold in there. The element ports can be made into ports for the loops (or the boiler, if desired)

Thanks for the tip, Mike.  By 'put the cold..." you mean the return side of the Hydraulic Separator?  With the 'hot' of the water tank/buffer connected to the boiler return?

Tom

He means the cold water inlet port to the hot water tank (which, yes, is what you would plumb to the return side of the hydraulic separator.) But I think he's also talking about making the TANK the hydraulic separator with a Tee at the top ( the hot-output) and a Tee at the bottom (the drain port.)

The cold feed to most water tanks is (for the plumber's convenience) located at the top of the tank not far from the hot-out port. To keep it the cold/hot from mixing or short-circuiting at the top of the tank, the cold water is directed to the bottom by an internal dip tube, which given enough time will corrode & fall off (much more quickly when used as a hot water heater than as a buffer tank, since fresh water is 100x more corrosive than "dead" heating system water.) By capping off the normal cold port to the tank and using drain port at the bottom instead, it takes the dip-tube out of the equation.

Whether plumbed in series with the boiler-return or Teed as a hydraulic separator, if it's easy enough to plumb it in that way it's a good approach.

Thanks Dana!

Yea, Dana speaks for me when I am mute....LOL

But the elements can be removed to gain two more ports as well.

A couple of last questions re using a water heater as a buffer.  

The Temperature/Pressure relief valve, that is typically mounted on the HWH, is not needed in this case (since there is one on the boiler), is it?

Also, is there a plug that will fit the heater element fittings so I can remove these and not worry about them corroding or leaking?

Tom

The T/P valve can stay where it is IF there is another properly sized one on the boiler, which there has to be or it won't get passed. The one on the tank is typically set at 150psi so it should never trip.

I wouldn't worry about the elements corroding and leaking. One reason they do corrode is that new oxygenated water is constantly being introduced and the tanks normal operation promotes galvanic corrosion. This is almost nil or at least far less when it is used as an un powered buffer tank.

I also don't think you need anything bigger than 60 gal, perhaps even 40gal but it depends on how you remove the heat from the tank, which zones are open or closed.

Personally, I try to stay away from a large number of zones unless the zoned areas are vastly different in their heating needs......such as a basement vs a sunroom.

Lastly, you will now have much more boiler water in the system so the exp tank will have to be upsized. I don't know the volume of the system but at minimum a #60 tank should be used, maybe a #90. Rule of thumb is that you can seldom oversize an expansion tank but it is easy to undersize one.

Sizing buffer and expansion tanks is a matter of established science and thus eliminates the need for any "rule of thumb". There are a few rules, such as volume, fill temperature, operating temperature and heat transfer fluid.

A 60 gallon buffer would be overkill, even with his current radiation & zoning (the very rough napkin-math on 40 gallons on his 10' baseboard zone appears in a prior post).

I was looking at 40 gallons as the probable lowest-cost/least work solution: Smaller electric water heaters are as expensive as 40 gallon versions (and sometimes more), so there's no need to fine tune it unless it simply won't fit in the allotted space. Without substantially more radiation or making it a single zone it'll need at least 20 gallons, but at 40 gallons it's already in no-brainer land no matter how it's zoned.

From these very useful discussions, I've concluded that a 40 Gallon tank would be the right choice IF I can get it into the crawlspace.

With the removal of all services in the crawlspace I was able to convert the full door and stairs down into a half bath and a small access closet. Since Hot Water Heaters generally don't bend, I'm going to have to revisit my model to make sure a 40 gallon will fit into the opening.

Surprisingly, I think a 48 inch tall HWH may fit better than a low boy because of the smaller diameter. If I find that a 40 doesn't fit, I'll have to try progressively smaller tanks until I can get one to fit. IF, as a possible consideration, only a 20 will fit, would TWO 20's in series (or parallel for that matter since it's all about thermal mass) work?

Tom

Posted By BadgerBoilerMN on 06 Sep 2012 07:18 AM
Sizing buffer and expansion tanks is a matter of established science and thus eliminates the need for any "rule of thumb". There are a few rules, such as volume, fill temperature, operating temperature and heat transfer fluid.

Badger, we are discussing the possible remedies for the Ops boiler issues. It is Ok to use a rule of thumb to give a general idea of what he should expect and that is all I have done. If he wants to pay an engineer to run through the numbers, it is up to him, but he may not.  If you want to do it for him, feel free. If he wants, I would be happy to provide the proper formulas, no skin off my nose.

I would never do a rule of thumb heat loss, nor do I size boilers or radiation this way but it is not unreasonable to give an opinion on the general direction he should take or the issues he will encounter. Stating that it is more dangerous to undersize a exp tank than to oversize it, and to give an example, should not be a problem. 

Posted By TomWS on 06 Sep 2012 01:16 PM
From these very useful discussions, I've concluded that a 40 Gallon tank would be the right choice IF I can get it into the crawlspace.

With the removal of all services in the crawlspace I was able to convert the full door and stairs down into a half bath and a small access closet. Since Hot Water Heaters generally don't bend, I'm going to have to revisit my model to make sure a 40 gallon will fit into the opening.

Surprisingly, I think a 48 inch tall HWH may fit better than a low boy because of the smaller diameter. If I find that a 40 doesn't fit, I'll have to try progressively smaller tanks until I can get one to fit. IF, as a possible consideration, only a 20 will fit, would TWO 20's in series (or parallel for that matter since it's all about thermal mass) work?

Tom

I'd add radiation  & combine zones before putting two 20 gallon tanks in series, but in practice 20 gallons may be enough even if you keep the same radiation & zoning- it kinda depends on the hysteresis around the output temp setpoint in the boiler's internal controls. 

With 20 gallons even on the zone with  10' stick o' fin-tube you're now looking at about 180lbs of water, and with ~30KBTU/hr (500BTU per minute) of excess output it's only ramping up at (500/180=) ~2.8F per minute.  I'd be shocked if you couldn't get at least something on the order of  1.5 minutes of burn time minimum out of it, even when serving just the single zone, with flows/delta-Ts properly tweaked so that it's running at min-modulation for most of the burn.

Do take the time least learn the napkin-math, eh?

Water at standard temperature & pressure has 8.34lbs/gallon (
It takes 1btu to raise 1lb of water 1F

3/4" plumbing has about 0.8lbs of water per foot 1/2" plumbing has about 0.4lbs of water per foot (< crudest possible generic numbers for easy math)

The volume of water in the boiler is usually specified (I didn't look it up, but YOU should.)

From there it's 6th grade math to convert BTUs/hour in to BTUs per minute, gallons per minute into pounds per hour, etc to calculate how fast the temp rises, and get a first-order estimate how long a minimum burn time at the boiler's minimum output might be at a given temperature hysteresis on that thermal mass.  I can't imagine any boiler would have as little hysteresis as 3F, but it might- most of the time it's not in the spec.  Figure on ~5F, give or take, but set it up and make some observations.

On a cold zone just calling for heat the initial burn will be a bit longer since there's a bigger delta-T on the water-mass, but after that it'll just cycle until the call for heat is satisfied.  Turn all of the T-stats for other zones off, crank the T-stat for the smallest zone up, then time the boiler's burns, and the time intervals between burns to get sense of what the minimum burn time will be.  None of your zones have enough radiation or mass to make a huge impact on burn time at low output temp, but there will be some.  Even if you crank them all on at the same time (or run it as a single zone), even at 140F you still have ~10KBTU/hr of excess boiler output than the radiation will deliver, so if your min-burn on the tiny zone was 1.5 minutes with 20 gallons of buffer, as a single zone the min-burn will still be under 5 minutes- it never balances for continuous burn.  And you'd need a LOT more radiation to get it to balance at 120F AWT.

The goal is to get the number of burns down to under 5 per hour under any conditions to minimize the wear & tear on the boiler.  Efficiency will increase if you can get the minimum burn time up, since every ignition cycle & flue purge throws away a fixed amount of heat. That wasted heat will be a significant fraction of the total burn on a 20 second burn, whereas on a 3-minute burn that loss-fraction is a bit more than 1/10th of that.  If with the lower mass you're looking at 6,8, even 10 burns/hour, it isn't the same disaster in terms of boiler wear that 30+ burns/hour would be.  But burn times of under a minute are an efficiency-disaster worth mitigating at current propane prices.

EXCELLENT reply, Dana!  Thank you very much!  This has helped me immensely!  Too bad this forum doesn't support 'likes'!

I've tried to 'install' a medium height 40 gallon tank in my 3D model and it is not clear it will fit.  A 40 gallon low boy or 30 gallon medium will.  I'll look at the information you've provided and decide if 30 Gallon is enough.

Thanks again,
Tom

it's pretty unlikely you'd need a 40 gallon buffer on even a midsize modcon. you only need to store about 5,000 BTUs per burn to get a 10 minute runtime out of a 30kBTU/hr min output, and that's with zero load factor. that's 600 gallon-degrees (5k/8.4 lbs per gallon). Divide by hysteresis and you have a buffer size. Hysteresis should be AT LEAST 20 degrees if you are trying to minimize cycling and I'd push to 30 if possible, but it depends on the max water temp you need and the max operating temp of the boiler. with wider settings you need to be careful you don't overlap room temp at the bottom end of your reset curve though or you'll have to wait for the house to get cold to fire the boiler again. ask me how I know this.... this is fixed by raising the start temperature of your reset curve.

and that is a 10 minute runtime, which is arbitrary and considered "ideal" by siggy. boiler MFGs who are paying attention to water content on mod/cons are still using much less and as Dana shows above it's probably overkill by a fairly wide margin.

When manufacturing the first condensing boiler, without the benefit of modulation, we considered burn over 5 minutes more than adequate. Buffer tanks were a common after-no-heat-load band-aid. Really no excuse then and even less now, but people will not change.

Posted By NRT.Rob on 08 Sep 2012 10:16 AM
it's pretty unlikely you'd need a 40 gallon buffer on even a midsize modcon. you only need to store about 5,000 BTUs per burn to get a 10 minute runtime out of a 30kBTU/hr min output, and that's with zero load factor. that's 600 gallon-degrees (5k/8.4 lbs per gallon). Divide by hysteresis and you have a buffer size. Hysteresis should be AT LEAST 20 degrees if you are trying to minimize cycling and I'd push to 30 if possible, but it depends on the max water temp you need and the max operating temp of the boiler. with wider settings you need to be careful you don't overlap room temp at the bottom end of your reset curve though or you'll have to wait for the house to get cold to fire the boiler again. ask me how I know this.... this is fixed by raising the start temperature of your reset curve.

and that is a 10 minute runtime, which is arbitrary and considered "ideal" by siggy. boiler MFGs who are paying attention to water content on mod/cons are still using much less and as Dana shows above it's probably overkill by a fairly wide margin.

Here is what I've measured/calculated:

The smallest zone is the Guest bedroom zone with 12 ft of baseboard and approximately 60 feet of 3/4" PEX (7.8 Gallon, including boiler's 0.95 Gal).

The boiler has minimum output spec'd at 32804BTU/hr.  Gallon Degrees = 655.5.

Running the system on this smallest zone, with outdoor temp at 66 degrees, the burn time was only 36-37 seconds with a temperature hysteresis of 16 degrees F (on at 33C, off at 42 C), setting the set point higher only shifted the on/off temps, the hysteresis was about the same.  Off time shifted, however, but that is sort of meaningless with these warm outside temps.

Dan, I didn't see any parameter that would change the hysteresis directly.  There was a parameter to shift the curve and the control on the front of the unit also does this.

Gallons required for a 10 minute burn: 41
Less Gallons in smallest loop: 33.2

Seems like a 30 gallon buffer would be 'ideal'.
A 20 Gallon buffer would be about a 6.8 minute burn.

Whatcha think?

Tom

Posted By Dana1 on 07 Sep 2012 12:08 PM <...snip>

3/4" plumbing has about 0.8lbs of water per foot 1/2" plumbing has about 0.4lbs of water per foot (< crudest possible generic numbers for easy math)

I'm struggling with 0.8 lbs/ft for 3/4" plumbing.  According to what I've seen, 3/4" plumbing would have about 2.8 Gallons per 100 ft.  This means 0.028 gallon per foot and, at 8.34 lbs/gal, that puts a 3/4" pipe at about 0.233 lbs per foot, not 0.8...

What am I missing?

Tom

Posted By TomWS on 09 Sep 2012 09:17 PM

Posted By Dana1 on 07 Sep 2012 12:08 PM <...snip>

3/4" plumbing has about 0.8lbs of water per foot 1/2" plumbing has about 0.4lbs of water per foot (< crudest possible generic numbers for easy math)

I'm struggling with 0.8 lbs/ft for 3/4" plumbing.  According to what I've seen, 3/4" plumbing would have about 2.8 Gallons per 100 ft.  This means 0.028 gallon per foot and, at 8.34 lbs/gal, that puts a 3/4" pipe at about 0.233 lbs per foot, not 0.8...

What am I missing?

Tom

Its effect may be minimal depending on how many feet of pipe there is. The tank and boiler piping may grossly over shadow the lack of volume in  the piping but you need to measure it first and know if it is type L or M copper (although the difference is minimal, 0.0251 vs 0.0268 gal/ft).

...and PEX has it's volume/ft that varies with temperature, etc, and they all have at least SOME thermal mass.

I did the fuzzy "in your head" math using "a bit shy of one water-gallon equivalent mass for every 50' " for generic 3/4 plumbing half that for half-inch and STILL got it wrong (by not doing the math correctly in my head- SHOOT ME! or buy me a cup o' coffee... :-) )

It's a rare system where even that gross error in the pipe-borne thermal mass makes a difference- in your system that error would be in the statistical noise of exactly how much water there ACTUALLY is in a nominal 40 gallon tank.

I didn't read the manual closely enough to see if the internal hysteresis around the outdoor reset setpoint is programmable in that boiler, but if it is you can get a lot out of even small buffers, as Rob described. A 16F hysteresis is on the high side compared to a few other non-adjustable boilers I've seen. My napkin-analysis was based on a hysteresis of 5F, which is lower than any boiler I've seen (but not a huge amount lower- I've seen 7F), a number chosen to demonstrated that it's a no-brainer bet. If you're measuring 16F of hysteresis, 20 gallons also enters no-brainer land- it's enough.

There's no magic in 10 minute minimum burns vs. 5, but if you're looking at a dozen burns/hour & counting it's boiler-abuse.

Burns under 1 minute in duration are just plain KILLING efficiency- as much as 10% of the source-fuel BTUs could be wasted on flue purges & ignition cycles on those 35-40 second burns, which gets reduced to a 1% source-fuel waste for a 350-400 second burn. A 10 minute minimum burn isn't dramatically more efficient than a 5 minute minimum burn.

Put another way, assuming the average return water temps would deliver 95% combustion efficiency steady-state, with the short-cycling boiler you're looking at about 85% as-used efficiency on the bursty 35 second burns, but with 20 gallons of buffer lengthening the burns by an order of magnitude you're getting ~94% as-used efficiency. Bumping it up to 40 gallons would buy you another 0.5%- there are diminishing returns.