Heat Loss Calculations

Balancing on one pipe steam isn't that hard -- and doesn't involve changing radiators around, which I wouldn't recommend unless you have some other very good reason for doing it.

There are two steps to it, though -- the first step is to make sure that the main venting is really adequate. If you haven't done that, you are pretty well doomed. Once you have that, though, the rest is easy enough, if a little tedious. You need less heat in a room? Slower vent. Slow things down, though, as much as you can before you start thinking about maybe putting a faster vent where you need more heat.

On steam boiler size: Steam boilers are not sized by the building heat loss, but rather by the installed radiation (if you were building new, you'd match the radiation to the space, then the boiler to the radiation -- but you're not).

foresthillsjd said:

@superj in that case, could the balancing problems be solved by slowing the venting on the closest radiators?

Balancing starts with venting the main very fast in proportion to the radiators. Then venting based on radiator size, but using small slow vents. You want to heat evenly. I’m starting to believe that anything larger than a Gorton or MoM 5 for a radiator with EDR under 80 is just a crutch for too little main venting unles your laterals or risers are very short or the boiler is oversized.

I underfired my boiler to a point where output it was just slightly smaller than the total radiation, so no pickup. No change in balance.

Jamie Hall said:

On steam boiler size: Steam boilers are not sized by the building heat loss, but rather by the installed radiation (if you were building new, you'd match the radiation to the space, then the boiler to the radiation -- but you're not).

Isn't she though? It's down to the studs, she did a heat loss calculation on each room (presumably using the the new insulation (and maybe windows) that she is likely installing, then ordered radiators to match.

True, the boiler would then be really oversized, but that's nothing new to anyone here, and if the existing radiation is too large (now very likely), slowing them down with small vents would have a similar effect on the system as having smaller radiators would. No?

PMJ said:

Ah, the endless debate over the size of the hardware.



The focus should be on the control to produce and deliver an amount of steam per hour that matches the current demand for heat. How big radiators are in the rooms makes no difference to the project if all there is for them to condense is the correct amount. How partially full the radiators get in condensing that steam makes no difference unless they are too small. The only place you go wrong is if the radiation in the room is too small and even a full radiator can't condense enough steam to meet the demand. Extra radiation hurts nothing if you control the amount of steam produced. Can't overshoot if there isn't the steam to overshoot with... no matter how big the radiator.

Ok,
You go ahead and get them a control to do that and let us know when it's on the market.

PMJ said:

@ChrisJ ,

There are many levels to that. The simplest one - a very simple duty cycle timer in series with the Tstat will do wonders compared to a vaporstat on a big boiler. I recommend people start there. I really don't think I am needed for something as simple as that. Somehow because no off the shelf control exists the whole control subject gets shelved and all the focus remains on hardware changes. I find that tragic really.

You already enjoy the benefits of a run time limiting control. A great approach isn't it.

I tried a simple duty cycle timer, it was absolutely awful on my system. Far worse than just the thermostat alone.

I'm very lucky to have a controller that had a huge amount of effort put into it and it works very well.

Most do not have this luxury, including the OP.

The OP is trying to redo their home, not engineer a boiler control.

If I am mistaken, and the OP actually wants to take on something like this on top of everything else they are dealing with, I am sorry, my mistake.

I think I should write an article..

First, on @foresthillsjd questions revolving around the existing boiler. It is not only wildly oversized (by a factor of 2) for the connected radiation, both the radiation and the boiler are oversized for the heat loss.

Having the radiation oversized relative to the heat loss is not uncommon. In fact, as the OP points out, even in situations where the radiation is sized closely to the design day heat loss, it will be wildly oversized most of the rest of the time. The OP's comments on the behavior of the radiation under various conditions are well thought out; it is, however, fundamentally a control problem.

Some systems (such as a mod/con hot water system) can modulate the output of the radiation to match the space heat loss. In that scenario, the radiation is always on, and always just warm enough to counter the heat loss. This, however, is not quite as simple to do as it sounds. The problem, of course, is that the heat loss is not governed solely by outside air temperature, which is the usual controlling parameter. Other factors come into play, such as solar gain and wind load and infiltration, as well as occupancy. Therefore such a system must, in fact, be slightly over designed, to ensure that in more extreme conditions (which can be either less or significantly more heat loss) there is adequate input -- which then must be managed. More on that in a moment.

Many types of systems cannot modulate the radiation output -- never mind the boiler (read on -- boilers are next) by varying instantaneous output. Instead, they must rely on varying the average output by having the radiation alternately heating up and cooling off. In control theory, this is referred to as pulse modulation, and there is nothing wrong with it -- properly managed. If one gets into it deeply, it can get rather complex, but in principle all that one is trying to do is to control the ratio of on time to the total time between individual on events, and to select that total time between on events so that the system performs as you want it to in other regards. Having elements in the controlled system which respond slowly to being turned on and turned off helps tremendously; one of the reasons large heavy cast iron radiators give such satisfactory results is the while they heat relatively rapidly, the cool much more slowly, evening out the heat delivery to the space while the heat source is off (in electronics, where pulse modulation is used all the time and is indeed inevitable, if the power supply is alternating current, this is done with capacitors)(one of the reasons why radiant floors of concrete can be so satisfactory is that the response times are really long -- on the order of hours to days)(piston engines are also pulse modulated machines -- only there the intensity of the on pulse is modulated, as well as the pulse frequency. The flywheel, of course, evens out the net power output).

The trick then is to control the on and off cycles, and it must be accepted that there will be some variation in heat input to the space during the cycle -- which means there will be some variation in temperature of the space over the span of a cycle. So the objective of the exercise is to keep that variation to a minimum (suggesting a high pulse frequency -- short cycle time) while ensuring that the system works efficiently (which usually means a somewhat longer cycle time -- long enough, for instance, for the heat supply to reach all of the radiation in the building).

There are various control devices and schemes to accomplish all of this. At the risk of sounding stone age, by far the simplest is a properly calibrated thermostat, located so that it reliably samples the space temperature. Some thermostats do better at this than others, primarily by adding into the space temperature a mechanism to compensate for the speed with which the radiation itself heats up and cools off. Some of the modern digital computer based ones can "learn" this behavior, and can do remarkably well. Some of the stone age ones, equipped with "anticipators" can do just was well, but do require an intelligent human to adjust them properly.

More, on boiler control, in the next post...

@Jamie Hall For the average person, I don't care if they use a thermostat or a vaporstat or a low pressure pressuretrol. As long as it's a well tested and reliable device.

For someone like @PMJ that has the ability and free time to make a better control, by all means go ahead, as long as they know what they are doing and the risks involved if the control should fail.

Most people do not have the ability to design and prove out a control.


@foresthillsjd If you're going to reduce the amount of radiation you have, I highly urge you to replace the boiler as well. A properly sized boiler will not only save you money in the long run, it'll run much nicer the entire time.

foresthillsjd said:

ChrisJ said:

@foresthillsjd If you're going to reduce the amount of radiation you have, I highly urge you to replace the boiler as well. A properly sized boiler will not only save you money in the long run, it'll run much nicer the entire time.

@ChrisJ, how much of an energy loss is this oversized boiler making? I spend about $570 a year on gas to heat the place, so a new boiler would have to cut the gas bill in half to make sense for me.

However, I did just learn today that I am going to have to replace my hot water heater. It's a 75 gallon tank. That thing is like a thousand bucks, and now I'm wondering if I should just convert to a condensing boiler, hot water heat that also can supply domestic hot water. I've fallen down this rabbit hole, deep.

rabbit hole

A steam boiler can also run an indirect for hot water.
@Steamhead knows a lot more about those setups than I do.

Personally I didn't do it because I didn't want my boiler hot all summer, but I can't say whether it was the best move when it comes to efficiency or not. I'm using a power vented 50gal tank heater.

$570 a year on gas isn't much and I doubt you could cut that in half by switching boilers.

Larry Weingarten said:

Hello, 75 gallons is a big heater. I'd start by sizing the tank to your actual need. Normally it's 20 gallons for the first person and 15 for each additional person, though with low flow fixtures you can do better. Large draws like a soaking tub mess with this approach. Once you have a tank size, then look at tying it into your boiler as an indirect (which other people here can do in their sleep) or possibly going with a heat pump unit... and/or waste heat recapture, like shower drain units, which can get 60% of the "waste" heat going down the drain.
There are lots of ways to reduce the load, so you can supply heat with a smaller, less expensive heater. I live down this rabbit hole!
Yours, Larry

While I like my 50 gallon tank, we actually did survive with a 120V 20 gallon electric heater for several months. And we were a family of 3 at the time. I used that temporarily while I put gas in the house. You had to plan around it to an extent, but it wasn't that bad really.

You know how things are now days though. Wasn't a 30 gallon gas heater considered normal years ago?

wow $570 for the year ?
i thought previously you had said it was double that..

i am in queens and payed about $2000 to heat 2200sq ft last year

Steam boiler sizing is according to the radiation installed and NOT a heat loss calculation!

Question: given that most early 1900's homes have much excess radiation installed (as in radiation that was never expected to be run full anyway); sizing to that will result in pretty big boilers relative to the actual design day demand won't it?

My radiation installed in 1926 doesn't need to be 1/2 full continuously to heat just fine at -20F. I don't think the boiler is really that much over being sized for that and needs to run about 1/2 the total time to get that heating done at -20F. Maybe this is why we have so many big boilers? If you size boilers to be capable of filling radiators that never need to be close to full to get the heating done anyway they will be...big.

Once you are getting steam to the radiators I don't think it matters how big they are as long as the boiler is big enough to heat the place. You only use as much of the radiator as is needed. Extra beyond that is extra. But if the boiler is easily big enough to cover the total heat loss that extra radiation isn't hurting anything. It doesn't somehow make the boiler too small. The only pipes that must heat up are those from the boiler to the rads.

Pay very close attention to @PMJ 's excellent point:"Once you are getting steam to the radiators...". That is really the fundamental key to the whole thing. I would at least partly agree that if your system can and does do that, and that the radiation is getting the steam it needs, you're good to go. Some systems -- like the one in the main place I care for -- need all of the installed radiation to handle the heat loss at and below the design day temperatures -- which we do get. Many systems where there have been extensive upgrading to the envelope don't.

But -- @PMJ 's comment is the joker in the deck. There are many variables involved here for a steam system (hot water, by comparison, is easy!): the size and length of the steam mains and risers, the way they are (or aren't) insulated, the physical arrangement of pipes and pipe junctions -- and of course the perennial favourite, venting (application of a variable vacuum comes under that heading, in my mind). The length of time between firing cycles comes into play, too. Each system will be different -- but the conservative approach of sizing the boiler output to the radiation demand with a safety factor is much less sensitive to such variations; even a system where the layout and execution out in the dark corners is, perhaps, less than optimal will work well -- or can be made to, with the major interventions being with regard to venting.

foresthillsjd said:

PMJ said:

Once you are getting steam to the radiators I don't think it matters how big they are as long as the boiler is big enough to heat the place. You only use as much of the radiator as is needed. Extra beyond that is extra. But if the boiler is easily big enough to cover the total heat loss that extra radiation isn't hurting anything. It doesn't somehow make the boiler too small. The only pipes that must heat up are those from the boiler to the rads.

So there's a nuance here that I want to pick at. I understand that if your goal is to heat the place, it doesn't matter how oversized the radiators are if the boiler is capable of handling the heat load. But IF the boiler is massively oversized for the heat load, would you be better off having radiators oversized for the heat loss as well, because then you would have less cycling? The boiler would get to run for a longer time to fill the radiator and then slowly cool. If the radiators were sized to match the heat loss, the boiler would be cycling more frequently for shorter intervals.

I disagree with sizing a boiler to the radiation, but most here disagree with me, so there's that.

An undersized steam boiler is difficult to balance, but if you take your time, it does seem possible. Of course, you should proceed at your own risk, knowing that it will be difficult.

In my experience, oversized radiation with a matched boiler will cause overheating of the home and is difficult to control with just a thermostat. Meaning with a digital thermostat for example, it will overshoot by a degree or two, and then stay off too long and then not heat enough the next cycle, and then overshoot again. Rinse and repeat.

My own system, I have 392 sqft of radiation in a 1600sqft house and I have a 104,000 btu/h output boiler rated for 325sqft assuming a 33% pickup factor.

So one of the main tenets to always size the boiler to the installed radiation is questionable, I'm happy to hear it. It's another weapon to use against oversized boilers.

ChrisJ said:

An undersized steam boiler is difficult to balance, but if you take your time, it does seem possible. Of course, you should proceed at your own risk, knowing that it will be difficult.

In my experience, oversized radiation with a matched boiler will cause overheating of the home and is difficult to control with just a thermostat. Meaning with a digital thermostat for example, it will overshoot by a degree or two, and then stay off too long and then not heat enough the next cycle, and then overshoot again. Rinse and repeat.

My own system, I have 392 sqft of radiation in a 1600sqft house and I have a 104,000 btu/h output boiler rated for 325sqft assuming a 33% pickup factor.

@ChrisJ thanks for posting your example. I think it's very pertinent.

@foresthillsjd If you have confidence in your distribution and venting, and the rads are similarly oversized to the heat loss on a room by room basis, it seems that tightening up the boiler sizing closer to the heat loss (maybe 130% of heat loss) would be the way to go. The exact size would be best determined by a steam pro who could examine your system in person and determine how aggressive you can be in tightening up the sizing. In the later TRV article I posted earlier, the author mentioned he has gotten pretty good at drawing a balance between sizing to the heat loss and radiation.

The extreme example would be a boiler sized to the heat loss heating a single massively oversized radiator. If the rad was oversized substantially it wouldn't matter since it would still condense all the steam (and release the BTU's to the space). As you add more rads it becomes more sensitive to distribution anomalies. But venting your distribution properly makes a complicated system behave more like the one rad scenario where all the rads get hot simultaneously even if they aren't saturated with steam. One rad may be more likely to become hotter than the other if it's easier for the steam to get to some rads, think long, uninsulated, exterior walls, undersized piping.

Jamie Hall said:

If you have perfectly even distribution, and if you have the radiation sized accurately to the heat loss, then I would agree that you can probably get away with a boiler sized to the heat loss plus a factor to get steam to the radiation which factor, if the pipes are short and well insulated, can be rather small.

If

You can then control the system to give just enough run time to get just enough steam to the radiators to get them just warm enough to counter the space heat loss on any given day. I'll leave the control philosophy to others -- but there are many possibilities.

Otherwise... the loss in efficiency from having a boiler correctly sized to the radiation so that, at maximum load, all the radiation gets hot and stays that way until the heating call ends is not that large; a few percent, perhaps. To the perfectionist, of course, this is not good. To the average client, and to the tradesperson who designs, specifies, and installs the system, maybe this isn't so bad. The client wants to be warm -- and to be able to warm up (people will set the thermostat back, or the power may go off...) and the last thing the tradesperson wants is to be called at oh dark hundred and 10 below with snow by a client who has a cold house.

As you already know, I believe in extra boiler anyway. I like to be able to cold start faster, and taking advantage of natural vacuum requires significant off time pauses - even on design day. And my abilities with special controls allow me to operate without temperature swings and pressure.

Having said that, what I have found in my own original 1926 installation is that installed radiation is well past what is really needed to cover the heat loss even at record cold. I believe this to have been the design for control reasons. A partially filled radiator is adjustable, a filled one not so much.

So I do agree with those like @ChrisJ who argue for using less pickup factor and therefore installing smaller boilers in conventional control settings. From the point of just what is required to heat, sizing to include condensing in all the radiation at once plus a pickup factor will result in a very large boiler relative to worst case actual demand.

I also think measuring % burn time of total time at design day cold demand will give a pretty good indication of where your boiler is vs heat load. Mine is at 50% at record cold. It is sized pretty well to the radiation - which even at record cold doesn't need to be very full to cover the demand...maybe 1/2 or so.

One other thing to consider. When we speak of how full the radiation is we must also talk about what % of the total time it is that full. Average fill level might be a better way. In the original design the fill level was quite constant as it was quite impossible to change it quickly anyway. With fixed 100% fire intermittent boilers the fill level required gets higher the fewer cycles there are. If you aren't going to fire very often you must fill the radiators fuller each time to get the average required. I chose to find a control that would be more like the continuous fire result at the rads. Obviously the fill level required to meet the demand is the smallest when it is constant. It only took me 2.5-3 cycles per hour with burns of 8-10 minutes each +/- to get rads that fluctuate between 180 down to 140 F along the top from the end of one burn to the start of the next. Calls for heat get longer and longer (into hours) the colder it gets. I really didn't like the cycles of boiling hot full rads with then long waits while they cooled all the way down. My heat is much more even now with vacuum continuing some flow going between the burns.