Gravity pipes as buffer

gasfolk

is in the center of the house. Moving the boiler to an outside wall (to facilitate thermal isolation) adds 15 feet to the exhaust (66 feet total developed length). Sidewall venting has significant complications. Cumulative costs start to become interesting.

gf

gasfolk

Short-cycling, gravity-conversion system...

Is it correct that the mass of water in a converted gravity system CAN act like a buffer, but only IF the supply temp is allowed a significant differential (e.g. 30F), which would require the system to be fully TRV'd to avoid marked swings in room temperatures.

OTOH, if the supply temp is closely controlled--with the boiler adding just enough heat to the match the heat loss--wouldn't the boiler short-cycle as badly as if there were no buffer effect??

Have I missed some key fact?

Thanks,

gf

gasfolk

The follow-up question is...

Either way, what control strategy might give the best advantage from such a buffer, and could it be used in series with an indirect/buffer tank?

Thanks,

gf

John Ketterman

For any specific on/off differential, more water means longer cycles. For any specific amount of water, larger differential means longer cycles.

You seem to think that it is necessary to have a lot of water AND a large differential or it's worse than having neither, but in fact, either one is good for avoiding short-cycling, and having both is even better. But as always, the best way is to not oversize the boiler.

I don't understand why you think a wide differential means TRVs are necessary. The thermostat will call for heat until satisfied, and during this time the aquastat will go on and off so the water temp will go up and down, but it will always heat the room towards the thermostat's set point. So there won't be any "marked swings in room temperature". Although TRVs are always a good thing.

A converted gravity system works better with constant circulation or at least delayed-off circulators. Otherwise, all that hot water sits in the pipes and releases heat where the heat isn't needed, like in the basement or in the walls.

gasfolk

Thanks for your comments...

Hi J. Cricket,

I see what you mean about large water volume and a large differential both being useful. Our gravity system is now piped Pri-Sec, with outdoor reset of the secondary (system) loop, so we have a lot of water, but its temperature is narrowly controlled by the reset curve (Tekmar 352). For example, on a recent design day, the supply temp was fairly steady at 125F (with boiler firing 4 minutes every 17 3/4 minutes).

Could we try for longer boiler burns by dumping Btus into the Secondary (system) supply pipes, (for example allowing the system supply to vary from 110F to 140F on a design day)?? The average supply temp would be the same, but would we feel some temp fluctuations in the rooms??

Is there a control that will handle this??

Thanks,

gf

John Ketterman

I can't figure out what you mean by "dumping BTUs into the secondary"...of course you need to put BTUs in the secondary or the house will be cold.

If the boiler is firing 4 min every 18 min, then the supply temperature cannot possibly remain steady at 125F when the boiler is off. Well, actually...the Tekmar you mention is a mixing valve controller, and maybe the temperature you are reading is the supply to the house and is holding at 125F because of variable mixing. Which is fine; you are still getting the benefit of the buffer formed by the water in the gravity pipes, via variable injection rate rather than variable secondary temperatures. If my guess is right, you need to look at the supply and return temperatures at the boiler (i.e. in the primary loop).

The Tekmar is already giving the boiler a differential between "on" and "off" temperatures. In fact, unlike a regular aquastat, the Tekmar intelligently varies the difference based on load, to reduce short-cycling.

If you are running only 4 minutes out of 18 on a design day, your basic problem is that your boiler is oversized, in fact it is humongous. It is OVER FOUR TIMES bigger than needed! No one can fix that problem other than a contractor carrying a new boiler.

gasfolk

Buffering...

Hi J. Cricket,

(quote) You are still getting the benefit of the buffer formed by the water in the gravity pipes, via variable injection rate rather than variable secondary temperatures.(quote) This touches on our question: can that buffering capacity be increased in a PRACTICAL way by varying the secondary temps--widening the differential of the supply?? Would living spaces overshoot and require TRVs?? Simpler if TRVs would be unnecessary.

In our system, most of the water is in one huge, single-zone, gravity-conversion loop, which gets Btus from the boiler through injection piping. Our expansion tank is in the sytem loop, and we regret the confusion in terms like "primary" and "secondary."

We're considering every avenue to improve efficiency. Before replacing the boiler, we'll correct the near-boiler piping and add a buffer tank. I was hoping that more effective use of buffering capacity in the gravity-conversion zone might substantially improve boiler efficiency (on an observational basis).

Thanks,

gf

Constantin

If you have fitted TRVs everywhere...

...then I see no reason not to use the large supply piping as a quasi buffer tank. Running the piping on a wide ΔT would then allow you to take advantage of the mass and hopefully reduce short cycling.

My only concern is whether the Tekmar will allow wide differential swings on the loop water temperature. After all, most OR controls are trained to maintain one loop temperature in response to outdoor temps. But then again, I am not familiar with your control and perhaps a call to hydronicsmike at Tekmar is in order. You couldn't talk to a nicer guy, BTW!

I hope that between the buffer tank and all those gallons in your supply piping that a way can be found to decrease the short-cycling you're experiencing.

gasfolk

No TRVs (yet)...

Hi Constantin!

and you're right, we'd have to ditch the 352. For us, the gravity-conversion buffer question is mostly theoretical. With available tanks and indirects, we can't imagine a practical application. Can you??

We're planning boiler replacement, but have venting issues: historic brick house; lots of windows limiting vent placement; persnickity neighbors who call the building inspector--so real issues for mod-con sidewall venting. If buffering improves part-load efficiency, we should get better sizing data and more comfort with that tiny load we calculated. When venting issues are clarified (that's another thread), perhaps a Vitodens 300, but probably not till 2007.

Best regards,

gf

Plumbob

The simplest way to increase the cycle time is to go into the Tekmar firmware and increase the differential setting. This is the difference between the boiler input on and off temperatures, not the supply temp to the house; the latter is held fixed.

gasfolk

Thanks, J. Cricket...

Your suggestion is a good one. Our boiler currently cycles on a Honeywell aquastat with 30F differential, and we may try Tekmar's Setpoint Control 150 for its 40F differential. Unfortunately, the 352 is an either-or control (for boiler reset OR mixing operation) with no boiler differential adjustment in mixing mode. With a design temp of 125F, we understand our current CI boiler wouldn't benefit from reset.

Since folks often mention the buffer capacity of piping in gravity-conversion systems, our curiosity in this thread was to understand whether one can maximize the practical benefits of that specific theory. Thanks for helping us do that.

gf

Ron Schroeder

Hi gf,

You might be able to keep the 352. If you are not using an aditional buffer tank, my first thought would be to pipe the heating zone as the primary and the boiler as the secondary. Put the 352 system temperature sensor in the Primary heating zone just after it's circ. The 352 control's Heat Relay would go to the boiler TT terminals to turn on the burner and the boiler's circ.

Unfortunatly, the 352 has only an automatic differential so I don't know how wide it will set itself.

Ron

gasfolk

BTW, Constantin...

When his schedule allows, our plumber and I will be making a pilgrimage to Warwick to take in the the sights. It's a long way to travel, but we hope worth the trip.

gf

Mike E_2

Condensing boilers don't have to be vented through a side wall. They can be vented vertically through the roof. Also, many areas allow the pvc piping to be run through the old chimney flue, as long as nothing else is using the flue.

Mike T., Swampeast MO

You [seem] to have said it perfectly Constantin--"I see no reason not to use the large supply piping as a quasi buffer tank". [emphasis added]

I certain seem to be getting such a "quasi buffer tank" with a gravity circulation system with TRVs and directly connected Vitodens.

That said, there is DEFINITELY a limit to that buffer effect. The evidence again comes from the difference between the temp the boiler reports and the temp that I can read on the surface of black iron about three feet away.

The buffer effect only has real meaning when heat loss is below minimum modulation. When load is below minimum output any boiler must operate with some temperature differential between cut-in and cut-out. The Vitodens can't break this rule, but I can't say that I've ever figured out just how it applies such...

What I can say about the buffer effect is this:

1) Boiler fires--if boiler temp rises above some threshold in the initial seconds of firing (before modulation begins), the cycle ends.

2) If firing continues, the boiler modulates to minimum (25 mbh). Temperature reported by the boiler rapidly climbs to target. Temperature I measure begins a much slower climb to target. The quasi-buffer is charging... This takes from about 30 to 90 minutes.

3) Boiler continues to fire as boiler temp slowly rises above target. With a good reset curve, the measured temp also exceeds target and when the measured temp equals the boiler temp firing stops.

4) After firing stops, both reported and measured supply temps begins a rapid fall in temp but return temp actually rises for some period after firing ends. The quasi-buffer is being consumed... This takes from about 5 to about 15 minutes. (Remember--it's now supplying the FULL system load but only had the difference of minimum modulation - load to charge.)

5) After this initial "buffer discharge", measured return and supply temps will be almost identical. Once that happens, the boiler fires--BUT almost always in a series of those very short firings that I call "pulses". While the pulses are occurring, the return temperature continues to fall. The "buffer" is still being consumed, but it has discharged to the point that it can no longer supply the full system load.

6) The "pulses" continue and become more frequent until the boiler no longer hits its' limit in the first few seconds of firing and the process repeats.

7) When system loss is very low (say 10 mbh) or less, the buffer never gets charged because the boiler always hits its' limit in those first few seconds. In these periods, measured supply temp always stays well below target.

Constantin

Yup...

I personally like running a stainless vent better, simply because nothing can really go wrong with them. Viessmann shows such a set up, with 60' allowable chimney length, as long as the boiler can derive its combustion air inside the boiler room.

Granted, this is not sealed combustion, i.e. not as ideal as one would like, but it is a workable solution. Our boiler room is sealed from the rest of the house and features a dedicated air supply. Seems to work like a charm.

I seem to recall seeing one condensing boiler that even allowed you to intake at ground and discharge through a chimney... (Baxi?)

gasfolk

Use of chimney flue...

would require minimum vent of 51 feet and loss of benefits of sealed combustion (reduced boiler contamination and improved efficiency). Not feasible to isolate the boiler room. Major obstacles to sidewall venting. So, looking doubtful for mod-con.

Thanks,

gf

Constantin

Hmmm...

... perhaps you can e-mail a diagram of your basement? I wonder how hard it would be to have a side-intake, and a chimney connection. Build a insulated closet around the mod/con and you have quasi-sealed combustion. That's what I did with my utility room...

Constantin

It's worth it!

If you can, schedule to meet with Jim McCarthy. Not only is he a great guy, but he also knows the Viessmann product inside and out. Plus, he seems to tolerate ambitious homeowners like us quite well. :-P

gasfolk

The boiler flue...

is in the center of the house. Moving the boiler to an outside wall (to facilitate thermal isolation) adds 15 feet to the exhaust (~66 feet total developed length). Sidewall venting raises significant issues for the neighbors and AHJ. Cumulative costs (perhaps and regrettably including the neighbor's legal assertions) start to become interesting.

gf

Guy_6

Venting

The vent length going up the chimney is not unreasonable, and therefore is no reason to toss that idea. The intake air can be drawn down the inside of the chimney, using the space around the PVC like a concentric vent. You would just have to be certain that the exhaust side terminates 12 inches above the top of the chimney to prevent cross contamination. The exhaust would still need to be in PVC all the way up.

Constantin

Good Points...

... perhaps building the closet at the center of the house would be a possbility?

These closets do not have to be big, you can always mount the balance of plant on the outside walls. Put an exterior-grade door on the the thing and a thick supply pipe of air from the outside... power-vent it if need be.

I'd call Jim McCarthy at Viessmann if that is the brand you're after and talk it through with him.