Gravity RE-Conversion---Any advice appreciated

With buffer injecting to the old boiler's top and drawing from the bottom, might thermal stratification (and gravity flow) be sustained? The pipes are huge, we've improved the shell, heat loss appears low, so might accumulating frictional losses be relatively small?

I've tried before to come up with a workable plan for such a forced/gravity hybrid. Problem is that you'd still have forced flow through the tank (or wherever you have a decent volume of water). Injecting to the top and drawing from the bottom would set up a counter-flow situation and I HIGHLY suspect that the forced injection would "win". You'd just be re-circulating the same BTUs through the injection loop with VERY few of them leaving the system via gravity flow. Injecting at the bottom and drawing from the top [might] be a better situation, but again I very much suspect that too many of the BTUs will just circulate through the injection loop. The flow will GREATLY favor the circulator path with its relatively high pressure differential! Don't forget that you'd need HUGE tappings for the gravity connections. A small bottleneck doesn't bother forced flow too much, but gravity isn't so forgiving.

CI radiator equations suggest no output below about 100F, yet our house heats fine at those temps.

Yours and MANY others. Those old tables are invalid at lower temperatures!

An equation estimating gravity flow may not exist but would be interesting. Any ideas where we might find one?

Such certainly exists! Write me with your fax number and I'll send that information from a relatively modern book. You'll be working in rather unusual measures including milinches (1,000ths of an inch) instead of FEET of head, and inches per second of velocity instead of FEET per second. You'll also be using pounds per hour water flow.

A typical two-floor gravity system generally operates with only about 2 INCHES of pressure differential on the TOP floor--about half that on the ground floor! This with a 30° delta-t. Lower delta-t means less differential. See why I suggest that a circulator anywhere in the system will cause problems?

4 x 2½" connections doesn't sound like much flow restriction to me... The boiler tappings may well be bushed down from 3" as well. Those are larger and more taps than available in any modern boiler of proper size. I seriously doubt any reasonably sized buffer tanks have such tappings either.

Hate to bring this up, but isn't your real problem an extremely oversized boiler? 2½ - 3x oversized comes to mind. Know you now have a buffer tank to try to address the problem. If anything gravity flow would probably make the oversizing worse because forced flow will move heat MUCH faster. Don't forget that your converted gravity system can likely still immediately absorb the energy of even a greatly oversized boiler and liberate it nearly as fast. Transmission and liberation of heat aren't your problem--you simply have WAY too much energy available even the the coldest weather and since the boiler can only fire or not fire, it would take an ENORMOUS buffer to effectively dampen that roaring beast.

Gravity mains [can] act as a sort of buffer, but the conditions are strict:

1) There must be continuous circulation.

2) That continuous circulation MUST BE SLOW--either via gravity or controlled via TRVs. A "wide open" gravity system with the typical B&G 100 moving 25 gpm or so has a dampening as opposed to buffering effect.

3) Even though the pipes are huge and hold quite a bit of water, the "buffer" does not have the buffering ability of a tank of the same size. The "buffer" will be fully charged VERY RAPIDLY with a non-modulating boiler.

In my system (with a Vitodens) the buffer only works within a limited range of system load. As far as I can tell, I get about 7,000 BTUs of "buffer" capacity with the "buffer effect" working in a 14,000 btu range. It ONLY works when system load (heat loss) is in the approximate range of 11,000 - 25,000. Above 25,000 the boiler modulates and the mains act as a damper instead of a buffer.

When heat loss drops minimum modulation, the mains begin to act as a buffer. The excess energy accumulates in the mains while the burner fires and is liberated after it stops firing. This is an inverse relationship--the more slowly it accumulates the more rapidly it is consumed.

Once heat loss drops below minimum modulation minus "buffer capacity" (25 - 7 = 18 mbh) the "buffer" begins to charge faster than the remaining output is being liberated by the radiators. Firing times become shorter and shorter. Once loss drops to minimum modulation - "buffer effect" (25 - 14 = 11 mbh), it seems to loose all effectiveness and the boiler operates in continual "pulse" mode delivering very short bursts of full output with the mains again acting as a damper.

*~/:)

things i have seen...

some poor people just lash things up and they work and have no clue whats what or the math or science behind any of it they just figure 'well, we need heat!'

hey and it works:) been working 30 years lashed into old gravity systems printed beacoup d years ago.

too lazy or not enough time to remove the old boiler they just tied right into it ...of course laying in all kinds of 1/2" pex was not on the agenda:) just something to think about...

Have still been pondering...

One SERIOUS concern with using the old boiler that way is that it will turn into the most efficient radiator in the house with the heat going right up the flue... I suppose you could disconnect and plug the boiler flue--then at least it will just be a radiator heating the basement...

While I could very well be wrong, I still suspect that the gravity flow will be disrupted when you move forced flow through the boiler. It might work and I guess you've lost little by trying.

You seem to be willing to try anything to avoid using TRVs. Their cost really isn't that high. Believe you have one or perhaps two 1 1/2" valves to contend with, but such really isn't that serious of a problem. I've yet to meet a gravity pipe that I cannot remove and I've honestly only crushed one pipe--that a 1" pipe with a 36" wrench and 48" cheater bar. That was a system with some sort of hardening, almost cement-like dope. Found a solution--repeated dousing with cheap (mainly petroleum distillate) silicon spray. You may even be able to avoid "shifting" the location of the radiator by using two reducing bushings at the radiator taps. (That presuming the rads have 2" taps--if they're 1 1/2" you need a bushing anyway.)

Please don't even think that you could use TRVs to eliminate overheating on selected rads using gravity flow.

Check the cv ratings of even the least restrictive TRVs against that chapter I faxed you and you'll see that they'll halt flow through the associated branches. The piping itself consumes most of the available force leaving almost nothing for restriction in the valve or the radiator.

*~/:)

bump

The only unconverted gravity system I've felt was slow, not particularly even and definitely inefficient (original coal boiler converted to gas). All the rest I've felt/worked on were converted to forced flow LONG ago.

"Slow response" was ALWAYS stated as a problem with gravity systems--this required that the fire be carefully tended, to include attempting to anticipate the weather. While a modern burner and thermostat helped with that problem, I really can't imagine that it does do efficiently.

That chapter I sent you regarding the mathematical design of gravity systems is relatively modern (1951). The earlier texts are different. While the more involved do touch on such design, they generally concentrate on the "valve area" design method.

"The delta T from radiator to room will be higher in a shaded, northern room than in a sunny, southern room; more Btus will hop off into the northern room, cooling the water more, and increasing gravity flow; fewer Btus will escape into the sunny room, maintaining radiator water temp, and reducing gravity flow?? Flows too low to measure, but perhaps enough to fine-tune a well planned system?"

There [seems] to be some serious problems with that theory. Yes, output will go down in a sunny room--but only after it is overheated! I cannot however imagine that those "unused" BTUs just magically hop into underheated spaces. Yes, output will go up in a cool room, but you've already assumed that its' underheated! The goal was to avoid such underheated areas to begin with!

Modern boilers simply are not designed for gravity flow so you have to go to extraordinary lengths to imitate an old boiler--effectively a big tank with HUGE openings out the top and near the bottom. Then you have to get heat into that tank in a way that won't disrupt gravity circulation. Honestly seems both impractical and inefficient to me...

I'd NEVER replace a gravity system--instead preferring to bring it into the modern age with a condensing/modulating boiler and TRVs. This produces a system that the Dead Men could only dream about. Plus, any imbalances caused by poor design, added insulation, roughened pipes, etc. will completely disappear.