Approaches to reducing boiler size

winnie · February 5

I know that a steam boiler must be sized to match the radiators.

My rough understanding is that each radiator has the capacity to condense a certain amount of steam, and if the boiler cannot supply enough steam for all of the radiators in a system, then some radiators will not get steam.

This pretty much means that in an old house, where the radiators were sized 'pre-insulation', you end up with a boiler and radiator system that provide too many BTU for the house. In my home (built 1935) on the coldest day the boiler will fire for less than 1/4 of the time.

My question: is there a reasonable way to reduce the EDR of the radiators, so that a smaller boiler can be used to better match the heat load of the home? For example, if you insulate a radiator do you reduce its EDR without causing other problems. Or are there other techniques that will reduce the amount of the radiator that actually gets used?

I'm not planning on changing anything; just curious on this point.

Thanks

Jonathan

EBEBRATT-Ed · February 6

2 pipe steam you can orifice the radiator inlets. One pipe steam you can't do too much.

If it were me, I would do a heat loss calculation room x room. This will tell you how much you are over radiated.

As you mentioned above normally, we are taught to add up the edr of all radiators, then the boiler MFGs add to that EDR a factor of 1.33 for a steam boiler for piping and pickup.

Many including me think this factor of 33% is too high. I would make sure any boiler installed covers the actual radiator EDR and then lower the PU factor depending on how over radiated you are .

Jamie Hall · February 6

As a practical matter… the easiest thing to do is just run the system for less time in any given hour… which you do most of the time anyway, since the system has to have enough capacity for the more extreme conditions.

There is almost no downside to that — very slightly less overall efficiency, but otherwise…

Sorry about being Mr. Obvious, but…

winnie · February 6

Fair enough; oversize the system and then run it part of the time. As you say, even a 'perfectly' sized system would run part of the time when the weather is warmer the coldest day you design for.

But how much is the actual efficiency impact of over-radiating the space, and then having a boiler oversized for the radiators?

My current estimate based on run time is that my house is over-radiated by a factor of 2 or 3, and the boiler is oversized by a factor of 2 for the radiators.

Thanks

Jonathan

Bernie_the_Brewer · February 6

Inevitably, some rooms will be warmer than others. You can put a blanket on the radiators in the warmest rooms to effectively reduce EDR for those rads.

Jamie Hall · February 6

I never said — and would never say — that the boiler should be oversized for the radiators. for a steam system In fact, I would nope that no responsible person would say that.

That it happens more or less routinely doesn't excuse it.

Let's try again. The boiler not ourput — most easily expressed in EDR — should never be significantly more than the installed radiation, also expressed in EDR. Because there is a built in safety factor in the EDR rating of the boiler, it can be less; with well insulated mains, as much as 20% less is usually satisfactory.

This is not rocket science. The job can be done in an afternoon with a table of radiator sizes and an adding machine.

Done correctly, you have a system which works together well. The boiler can reliably and evenly power the radiators. You can then control that with a thermostat so that the power delivered to the space by the system matches the power required to keep the house at the desired temperature — also evenly and reliably.

Provided the temperature swing from the thermostat — not the boiler or radiators — is kept reasonable, the efficiency impact from turning the system on and off, say once an hour, is minimal.

The efficiency impact of an oversize boiler in comparison to the radiation can, however, be significant, depending on just how bad the oversize is and how the boiler is controlled.

dabrakeman · February 6

IF you are over radiated the you just won't ever completely fill the radiators on a balanced system. There is no issue with that. I.e., the system wouldn't know if you just cut 1/3 of the sections off each radiator. This is different than being properly radiated for the space but having an oversized boiler which can be problematic.

jesmed1 · February 6

We're in a similar situation with two WGO-5 hot water boilers in a 100-year-old 4-unit condo building near Boston, and all original cast iron radiators. The boilers run about 25% of the time on a design day, similar to yours. Your boiler is making steam; ours is making hot water. Your boiler concentrates its heat into a few gallons of water to produce steam; ours distribute their heat into 100+ gallons of water each to produce lower-temp hot water. But both our systems run about 1/4 duty cycles on cold days.

Since these are non-condensing boilers running at 80%+ "dry gas" efficiency, the loss of almost 20% of the heat in the "dry gas" exhaust, plus another 7% or so latent heat of vapor loss (for oil) and 15% or so latent heat of vapor (for gas) is already "baked in" to the boiler, and those losses can't be recovered without changing horses to a condensing boiler with lower exhaust gas temps.

So those of us with non-condensing boilers are left with trying to make marginal improvements in overall system efficiency. For example, recovering some of the heat stored in the boiler's cast iron heat exchanger after shutdown. In my case, someof that heat gets recovered during a typical 2+ hour idle time, during which I have our circulators set up for a thermal post-purge until the supply temp drops to 90 F. But even without the post-purge, gravity circulation would continue to suck some of that residual heat out of the boilers. I've done some calculations, and I figure that there's at most maybe 10% of a typical cycle's (45 minute boiler run) heat left in the boiler water and heat exchanger. We recover some of that , but some of it goes up the chimney as air continues to draft through the boiler after shutdown. So maybe we lose 5% up the chimney and recover 5% in post-purge circulation.

Some people try to retain that last 5%-10% of residual heat by installing automatic vent dampers. The downside of those is that when the damper motor fails and the damper won't open, the failsafe switch prevents the burner from firing, and you have no heat. I've decided the marginal possible gain isn't worth the hassle.

But back to your original question, the reason boiler oversizing matters is that a larger boiler has more thermal mass in the heat exchanger and in the higher volume of water in the boiler, both of which increase the amount of stored heat that can be lost up the flue after boiler shutdown. If we had boilers that were, say, half the mass of our WGO-5, we'd have roughly half the number of BTU's stored in it after a cycle, and thus roughly half as much heat to lose up the flue. So maybe we'd go from losing, say, 10,000 BTU of a cycle's heat production up the flue post-shutdown to maybe 5,000 BTU, or a 5% overall efficiency improvement.

Our boilers run about 500 cycles per season, so 5,000 BTU saved per cycle would be 2,500,000 BTU, or about 18 gallons of oil (out of 600 gallons burned per boiler). That turns out to be about a 3% savings. It's something, but not much.

The numbers above are ballpark-ish, as there's no easy way to measure how much heat we lose up the flue after boiler shutdown. But based on the above, I'd be surprised if we saved more than 10% fuel by going to smaller boilers (even though I plan to do so when these fail).

winnie · February 6

Hmm. Ok, I started this thread mainly out of curiosity, but now it seems I might have something actionable to save fuel.

I have an oil fired single pipe steam system. It is reasonably well balanced with even temperatures throughout the house, and all the steam valves work properly. The mains in the basement are insulated with 1" fiberglass, the risers in the walls have the original asbestos insulation.

Last night I measured my radiators, and have 245 square feet of of EDR. The label on my boiler says:

I=B=R Input 1.20 USGPH

D.O.E. Heating Capacity 145,000 Btu/hr

Net I=B=R Output Steam 108 MBH 450 SQ. FT

I have an hour meter measuring the run time of the burner blower, and use 1.05-1.1 GPH averaged over a heating season.

Is this boiler oversized enough that it will make a significant difference to fuel consumption? The system seems to run just fine.

Thanks

Jonathan

jesmed1 · February 6

@winnie said "Is this boiler oversized enough that it will make a significant difference to fuel consumption? The system seems to run just fine."

You may have missed my previous post before asking this question. My short answer is that, after a lot of data collection and calculation, I don't expect more that 10% fuel savings from reducing our boiler size/capacity by half. I'd be pleasantly surprised by better savings, but I don't think it's going to happen.

Jamie Hall · February 6

Your boiler is oversized enough (I wish people wouldn't do that, but…) that you may be able to improve things slightly. Oddly enough n(or maybe not!) the thing I would try first is the settings on the pressure control. You don't mention anywhere what your pressure control device is — or what pressures it is set at — but if this is a conventional two pipe steam system, or a one pipe system, set the cutin pressure at 0.5 psig and the cutout at 2.0 psig, and add a low pressure gauge if you can to verify those pressures. Various pressure control devices differ in how those two settings are made; if you could post a photograph of the device we can help with how to do this. If this is a two pipe vapour system, the two settings are different: cutin at 0.2 psig and cutout at 0.5 psig.

What is the objective here? The boiler is oversized for the radiation, so it is necessary for it to chut down from time to time to let the radiation "catch up". Th e shorter that shutdown time is, the greater your overall efficiency will be.

The second step is the thermostat. If it is a digital modern one, you may have better results setting it for 2 cycles per hour rather than one. Sometimes this helps. Sometimes not!

winnie · February 6

Thank you for the suggestions. I will get the information about the pressure control on Sunday and the thermostat on Sunday.

-Jonathan

PMJ · February 8

@winnie , all of the houses built pushing 100 years ago with steam heat had extra radiation installed that was never intended to be filled. The steam was supplied by a continuously fired boiler that never ran on "high" like ours do. The systems were designed to have a big boiler operate in the middle of its range where it could be adjusted, partially filling radiation and adjusting that steam amount to the demand in real time. So adding up all the radiation and sizing a boiler based on that results in boilers that if only running on high will quickly overfill the radiators for the demand.

The unfilled part of radiators downstream of what is needed for the current demand has no effect on anything. No modification to that would be required to change to a smaller boiler. How much radiation gets used is controlled automatically by how much steam you put into it.

This is why the standard setup with a boiler sized to the installed radiation(or in your case even more) running on high until the thermostat is satisfied is often so unpleasant. It overfills the radiation, way past what would be needed for the average day demand if supplied continuously. When the tstat finally reacts and shuts the boiler down there is an unfortunately long wait period while all that supplied heat dissipates. The boiler and mains fill with air and cool more than they need to resulting in a longer reheat period on the next burn and longer time to steam. It is a roller coaster ride that should have much smaller hills and valleys.

What is needed is simply to spread out the required burn time per hour of whatever boiler is installed (in your case 15 minutes per hour on design day) into roughly two evenly spaced burns. Yours would end up probably 8-9 or so minutes each in the very cold. A way to facilitate longer colder start burns in moderate weather minutes also very helpful. The start/stop/length of the burns needs to react to the actual sensing of the steam level out in the system and adjust itself. A vent damper is absolutely required, it just can't sit on top of the boiler if you want it to last. It will keep the boiler much warmer between burns reducing time to steam on every burn. Mine is 30 years old and I have never had an issue with it.

Though hardly that complicated, no off the shelf control was ever accepted by the powers that be in steam to do these simple things at the residential level. So contractors have nothing to offer. So I did them myself many years ago. I also resurrected the lost concept of simple natural vacuum between the cycles which has a very significant evening out effect which is not well understood. Between the two things the result is dramatically more even and certainly more efficient heating.

You appeared to be mostly on an intellectual exercise to understand more about what you are observing. Hope this helps.

Jamie Hall · February 8

I must have some sort of weird setup. I have a mercury T87 thermostat. Vapourstat. 2 LWCOs. More radiation than needed for the house (although at -12 this morning… and the boiler sized to fit the radiation..

No other controls.

No insulation to speak of (house built from 1780 to 1893)

And the building, at the thermostat, stays within half a degree either way of the setpoint, in design cold or fall and spring.

How odd…

PMJ · February 8

@Jamie Hall , I think we've been around this block before. We understand that you are quite happy with your system and control and that you don't believe any improvement is possible, needed, or worth the effort , one or all three. Please rest assured that your position is really fine with everyone as far as I know. I do not write here with any interest in getting you to change it. I write for those who just might be interested in some improvement, not to debate with anyone what is good enough.

The thing is, I have experience with both systems. I started out 35 years ago with exactly yours. I had a mercury T87, a vaporstat, and a boiler sized to the radiation (1000 edr), and a house with no insulation. Was it ok? I guess so as it satisfied all the folks who lived there for 36 years before me and intermittent fire. No one changed anything. True enough, it couldn't be that bad.

But as an engineer I thought it could be a lot better and I was right. I realized that by definition, the more even the heat is the longer the calls would be - assuming the boiler is actually always providing enough net steam to raise the temperature. As we all must know, if the temperature was perfectly even at the thermostat, no cut in or cut out would ever occur and calls for heat would be endless. So, again by definition, calls for heat that always end with the boiler running on high, or, having been going on and off during the call due to pressure stops, will be shorter with more temperature variation during them per unit time than much longer calls than terminate unrelated to whether the boiler is running or not. This is not a matter of opinion, it is by definition. I also realized that any pressure was a sign of too much steam in the system and always a negative and that there really should be no need for a vaporstat, even with intermittent fire.

So now I am standing in what appears to be the unique position of actually having lived with and operated both systems and therefore able to comment on the relative performance of them.

Jamie Hall · February 8

If it's still working 100 years from now, I'll take it. Well, 96. Cedric's system's birthday.

Captain Who · February 8

There's ambient room temperature fluctuation in the t-stat room and then there is radiator temperature fluctuation. The key is to have the minimum radiator temperature fluctuation because human beings are not t-stats. We feel the radiation of the radiator along with the convection of the ambient air and having that radiator get cooler in between heating cycles is simply never going to be as comfortable. Both situations could have an ambient air temp fluctuation of only 1/2 degree but will not feel the same.

PMJ · February 8

This year my system piping turns 100 and my boiler is at 69 years.

Jamie Hall · February 8

but the controls?

I just prefer simplicity and reliability. Don't mind me.

On another note… I have three trucks and two cars. It was -12 this mroning. One of the cars — a hybrid — started. The other, 10 years old, no hope. One of the trucks — 8 years old — no hope. The other two — one 53 years old, the other 30 — complained bitterly… and started; neither of them had been run since October.

Approaches to reducing boiler size

Comments

Categories