Boiler vs radiation sizing

ChrisJ

So something just occurred to me.

This may apply to 2 pipe systems as well, however I have zero experience with them so I'm mainly talking about single pipe right now.

I've mentioned before that I disagree with those who say we need to heat all of the piping and radiation together. This is simply because we do not, we first heat all of our piping and then our radiation starts getting steam as long as our venting is all correct.

Now picture heating the radiators, and bear with me.
Once all of the piping is at steam temperature the steam starts to spill into the first section of every radiator. It then fills that section for the most part and starts filling the next, rinse and repeat.

It's true, it usually will go across the top faster and other oddities but overall it slowly fills one section at a time.

So, we do not fill entire radiators all at once either.

With all of that out there.

What's the difference with a house that has all 5 section radiators and a house with the same model radiator, but all are 10, or 20 sections? The EDR goes way up but each section is still the same. So as the system is heating, up to 5 sections the boiler sized to the first house should heat all of the radiators in the other houses at least up to 5 sections fine. And since we heat one section at a time, this shouldn't be a problem.

Thoughts?

Gsmith

@ChrisJ, so the nub of your question is really why do we size steam boilers to the radiation rather than to the heat loss of the house? And you example is in fact what I believe happens to single pipe steam systems with appropriate sized boilers on days when the outside temperature is above the design day temperature. Indirectly, the boilers are sized to the heat loss, at least at the time the house was designed and built, because the radiators in each room were presumably sized to the approximate heat loss in each room or space. Of course, most houses have had changes made which alters the original heat loss (insulation added, infiltration sealed, better windows, etc.) so you example is probably more often the case than not--the boilers are over-sized. I'm certainly no expert, but in my experience with just my own 1920's wood frame one-pipe steam system, I think the boiler size mostly matters in terms of how cold a day can the house be comfortable in. If the radiatiors were sized closely to the heat loss in each space and not too many changes were made, or the changes to the spaces' heat loss were more or less equal in all the spaces heated, and the thermostat is appropriately situated, then the boiler should respond to the thermostat and not shut off on pressure except perhaps when coming back from a setback, even on days warmer than design day. I basically agree with your example, the radiators will all heat partially and proportionately, the spaces are warmed and the thermostat controls the boiler. And, I think in most cases the boilers used can heat the structure adequately at lower than design day temperatures because of the improvements made over the years that reduces the structures heat loss.

So, I don't particurlarly see why one couldn't size the boiler to the structure heat loss, as long as (important), the radiation is sized proportionately to each spaces' heat loss. If the structure is over-radiated (because of conservative design or heat loss improvements) you have essentially a system designed for a lower design day, and if you size the boiler to the radiation you may also be sizing the boiler to a lower design day temperature if the heat loss of the structure is currently lower than the original heat loss because of improvements.

Is that what you were getting at?

PMJ

@ChrisJ ,

I agree with you and I've always thought this. I don't think unused radiation downstream of that needed to heat has anything to do with anything - one pipe or two. This is one of the things that becomes pretty obvious when you pace a boiler and only fill to a level required to heat. Pressure stop controls fill everything up sort of making it appear that is required somehow. It really isn't. So a boiler matched to installed radiation will be bigger than actually needed. I like a lot of extra boiler for other reasons but it is not required.

As I've said before I think most of us have radiation well beyond what was ever required to heat on design day not because we have added insulation but because running continuously the rads could not be full and have any chance for the ultra low pressure damper controls to work well. I think the rads were designed that way to be significantly less than full even on design day. Mine clearly are significantly more than needed and I haven't added any insulation.

As long as there are enough sections available in a room to heat on design day connected to a boiler big enough to provide that much steam, I don't think having extra sections available beyond that affects anything except space and installation cost.

retiredguy

Let me tell you what I witnessed at a nursing home in Indiana. Pa. many years ago. The building envelope was updated with new thermopane windows from single pane, the roof was replaced and insulation was increased to the standard of that day, and the walls were insulated with some amount of insulation when they had none before. Every thing should have been great but the Engineering firm that was supposed to engineer the heating system and the new steam boilers sized the boilers (2 of them) to each produce about 60% to 70% of the total building load. There was nothing changed inside the building; no rad changes, no piping up grades, nothing. The returns were shot so there was almost no condensate return. The boilers ran on mostly fresh and untreated water which greatly increased the heating load. When the boilers fired the rooms near the boiler room would greatly overheat and the rooms far from the boiler would receive almost no heat since the boilers were controlled by a simple thermostat located near the boiler room. I only mention this because if the steam rads had been resized to the new heat loss every thing would have been OK and the boilers would have done a good job. I told the representative of the Engineering company that they screwed up I got a phone call from the owner of my company to get an **** chewing and was told that it was not my place to talk to them and to apologize. There were so many other problems with this job .

ChrisJ

retiredguy said:

Let me tell you what I witnessed at a nursing home in Indiana. Pa. many years ago. The building envelope was updated with new thermopane windows from single pane, the roof was replaced and insulation was increased to the standard of that day, and the walls were insulated with some amount of insulation when they had none before. Every thing should have been great but the Engineering firm that was supposed to engineer the heating system and the new steam boilers sized the boilers (2 of them) to each produce about 60% to 70% of the total building load. There was nothing changed inside the building; no rad changes, no piping up grades, nothing. The returns were shot so there was almost no condensate return. The boilers ran on mostly fresh and untreated water which greatly increased the heating load. When the boilers fired the rooms near the boiler room would greatly overheat and the rooms far from the boiler would receive almost no heat since the boilers were controlled by a simple thermostat located near the boiler room. I only mention this because if the steam rads had been resized to the new heat loss every thing would have been OK and the boilers would have done a good job. I told the representative of the Engineering company that they screwed up I got a phone call from the owner of my company to get an **** chewing and was told that it was not my place to talk to them and to apologize. There were so many other problems with this job .

If radiators near the thermostat were getting steam faster than ones further away that sounds like a balancing issue to me.

Was this a one pipe or two pipe system?

Why weren't returns replaced? A boiler won't last long under such conditions.

pecmsg

retiredguy said:

Let me tell you what I witnessed at a nursing home in Indiana. Pa. many years ago. The building envelope was updated with new thermopane windows from single pane, the roof was replaced and insulation was increased to the standard of that day, and the walls were insulated with some amount of insulation when they had none before. Every thing should have been great but the Engineering firm that was supposed to engineer the heating system and the new steam boilers sized the boilers (2 of them) to each produce about 60% to 70% of the total building load. There was nothing changed inside the building; no rad changes, no piping up grades, nothing. The returns were shot so there was almost no condensate return. The boilers ran on mostly fresh and untreated water which greatly increased the heating load. When the boilers fired the rooms near the boiler room would greatly overheat and the rooms far from the boiler would receive almost no heat since the boilers were controlled by a simple thermostat located near the boiler room. I only mention this because if the steam rads had been resized to the new heat loss every thing would have been OK and the boilers would have done a good job. I told the representative of the Engineering company that they screwed up I got a phone call from the owner of my company to get an **** chewing and was told that it was not my place to talk to them and to apologize. There were so many other problems with this job .

I've got 2 words for alleged Engineers and there not Happy Birthday!

Jamie Hall

In a way, very good points. And on anything but a full load design day, quite accurate -- and in a perfect world the boiler would modulate to produce just enough steam to fill only enough of the sections of the radiators which were needed to meet the heat loss of the space.

Leaving aside for the moment the somewhat vexed question of how to modulate a boiler -- since we have seen a wide range of ideas on the subject, most of them quite valid -- one way to do this would, clearly, be modulating valves on the inlets to all the radiators which were sensitive to the space temperatures served. Modulating TRVs, in short. Then all the boiler controls would have to do is modulate the boiler output (here we go again) to match the steam demand.

Instead, we hopefully balance the radiators and manage the venting so we can do much the same thing with a single control -- the thermostat.

But...

That's anything but a design day (or worse) which, granted, are few and far between. On such a day, though, if the system as a whole has been properly designed for the building, and the building hasn't been changed much since - both pretty big ifs -- we make an interesting discovery: the boiler is going full song and all the radiators are hot all the way across -- and the building is holding temperature. In short, for that odd day, we need every bit of that boiler capacity to feed the radiation to meet the heat loss.

I would make an analogy to aviation: I have two (or 1 or 3 o4 4 or 6!) big engines on my bird. Do I need all that power to stay in the air? Nope. a Boeing 777 can cruise for more than 2 hours on only one engine, for example. 747s have crossed oceans on three out of four. So why have two? In the case of aviation, I need both of those things out there, pretty close to wide open, just to get off the ground.

Or my friend's Peterbilt. Does he really need 700 plus Cummins ponies as he cruises down the Interstate in Iowa? No way. When he gets to Colorado heading west from Denver, he wants every single one of them so some swift kid on a bicycle doesn't tailgate him on the way up the hill.

PMJ

retiredguy said:

Let me tell you what I witnessed at a nursing home in Indiana. Pa. many years ago. The building envelope was updated with new thermopane windows from single pane, the roof was replaced and insulation was increased to the standard of that day, and the walls were insulated with some amount of insulation when they had none before. Every thing should have been great but the Engineering firm that was supposed to engineer the heating system and the new steam boilers sized the boilers (2 of them) to each produce about 60% to 70% of the total building load. There was nothing changed inside the building; no rad changes, no piping up grades, nothing. The returns were shot so there was almost no condensate return. The boilers ran on mostly fresh and untreated water which greatly increased the heating load. When the boilers fired the rooms near the boiler room would greatly overheat and the rooms far from the boiler would receive almost no heat since the boilers were controlled by a simple thermostat located near the boiler room. I only mention this because if the steam rads had been resized to the new heat loss every thing would have been OK and the boilers would have done a good job. I told the representative of the Engineering company that they screwed up I got a phone call from the owner of my company to get an **** chewing and was told that it was not my place to talk to them and to apologize. There were so many other problems with this job .

Rads could only be sized to max possible heat loss. All the rest of the time they will be physically too big for the heat loss - on an average day at least 2 times too big(no getting around that). @chrisj is right here - proper balancing is required to produce a relatively even fill everywhere at the same time or there is no chance at success.

PMJ

"That's anything but a design day (or worse) which, granted, are few and far between. On such a day, though, if the system as a whole has been properly designed for the building, and the building hasn't been changed much since - both pretty big ifs -- we make an interesting discovery: the boiler is going full song and all the radiators are hot all the way across -- and the building is holding temperature. In short, for that odd day, we need every bit of that boiler capacity to feed the radiation to meet the heat loss."

@Jamie Hall , my observation is that they installed radiation well past what was needed design day because a continuously firing system could not be controlled with rads filled and the pressure going suddenly higher. So rads were installed that literally never needed to be full...mine don't - even with intermittent operation.

The way most run today going totally full to pressure stop and then long waits in between the rads would need to get some fuller during the time steam was actually being produced. But the truth is the more evenly you supply the steam and the more continuously you do so, the less radiation is needed to get the job done.

But I think the question being discussed here is only whether a bigger rad beyond what is needed for the conditions changes anything. In other words can rads really be too big for a boiler truly able to satisfy the actual demand.

Jamie Hall

Actually, @PMJ , I agree with you. There is one critical sentence in my comments: if the system was correctly designed for the building... which as you point out, correctly, it often wasn't. Overkill was just as prevalent 120 years ago as it is now. Worse, many (not all) buildings have been substantially changed since then. This very often results in the building being over-radiated, sometimes by a very wide margin. This, I think, is the root cause of many of the balancing and control problems we face more or less daily.

I happen to be blessed with caring for buildings where this hasn't happened -- one steam, one hot water, one forced air -- so I'm probably a little less sensitive to some of these problems. All three of them are capable of maintaining comfort on the coldest days we get around here -- but only just barely.

PMJ

Jamie Hall said:

Actually, @PMJ , I agree with you. There is one critical sentence in my comments: if the system was correctly designed for the building... which as you point out, correctly, it often wasn't. Overkill was just as prevalent 120 years ago as it is now. Worse, many (not all) buildings have been substantially changed since then. This very often results in the building being over-radiated, sometimes by a very wide margin. This, I think, is the root cause of many of the balancing and control problems we face more or less daily.

I happen to be blessed with caring for buildings where this hasn't happened -- one steam, one hot water, one forced air -- so I'm probably a little less sensitive to some of these problems. All three of them are capable of maintaining comfort on the coldest days we get around here -- but only just barely.

But I didn't think I said extra radiation was not correctly designed. I don't think a continuously partially filled radiator getting the job done on design day(or worse) is over-radiated. I think they did that on purpose. But today, when boilers on one speed (high) are allowed to fill up that same rad(something that is never needed), it is suddenly deemed too big and a design flaw. This is the whole basis of my pacing or PWM work. Just a little pacing (which a pressure based control can't do) allows for operation at much lower pressures, never filled radiators, and much more even heat...much more like what was originally designed.

pecmsg

Jamie Hall said:

Actually, @PMJ , I agree with you. There is one critical sentence in my comments: if the system was correctly designed for the building... which as you point out, correctly, it often wasn't. Overkill was just as prevalent 120 years ago as it is now. Worse, many (not all) buildings have been substantially changed since then. This very often results in the building being over-radiated, sometimes by a very wide margin. This, I think, is the root cause of many of the balancing and control problems we face more or less daily.

I happen to be blessed with caring for buildings where this hasn't happened -- one steam, one hot water, one forced air -- so I'm probably a little less sensitive to some of these problems. All three of them are capable of maintaining comfort on the coldest days we get around here -- but only just barely.

@Jamie Hall
That's exactly my point!

The dead men were under similar circumstances…...We don't want to be cold! Things were oversized but on a smaller scale then today. Now that smaller scale 100 years ago with todays insulation, thermopane windows, Spray foam......Now the heat out put / Radiators are seriously oversized!

PMJ

Ok so I'll try this again. I think @ChrisJ gets it as it is why he started this thread.

Partially filled radiators are easier to control in all conditions. Radiators can only be one size. They must be big enough for design day. Therefore they must be "oversize" for all other days for all those who insist. I think worrying about the size of radiators is a fool's errand. It is all about the control. The original control ran along continuously with a boiler at 1/2 speed +/- and partly filled radiators in all conditions. Then along came on/off control. Someone decided to let a boiler making steam at design day rate run until either the tstat was satisfied or all the rads were filled and a pressure stop was hit. Dumb idea. In these systems that means big radiators overfilled for the conditions....basically every day. But instead of changing the control people only seem lament that they have the wrong hardware. But there is no filled radiator size right for all conditions darn it! When you get that you realize a big radiator can be right for all conditions....partially filled. It takes a better control like Ecosteam to partially fill them appropriately for the conditions.

Jamie Hall

Um. Well... um. I would argue -- granted, off rather limited real life experience, that a properly sized (that is, boiler matching radiation, radiation matching heat loss load under effectively maximum conditions) does not mean that the radiation is oversized (by definition it matches the effective heat loss) nor that the boiler is oversized (it matches the capability of the radiation to condense steam).

And I will go so far as to say that any other situation is improperly designed. To put it another way, if the system will not meet the spec. -- comfort requirements -- under design conditions, it is a failure. If portions of it are significantly overdesigned, that is poor design -- although at least it isn't an outright failure.

Now I will converge back with @PMJ 's thinking that the problem -- or challenge -- becomes one of proper control of the available hardware for off-peak operation. First, the radiation will, if properly controlled, be only partially filled under off-peak conditions. The challenge there, as we have all seen, is to get even distribution of steam; although radiation is self-balancing to a very limited extent, that still means that however it is done the proper amount of steam is sent to each radiator. It can be done more importantly, it can be done entirely passively. No active controls are needed. Second, the boiler will clearly not be operated at full capacity.

How is this to be done? a bit of history: many coal-fired boilers in the early days were actually modulating boilers. There were a variety of fascinating contraptions which accomplished this -- and most of them worked pretty well so far as controlling the output. A few of these used space temperature directly as the control (one of the stately homes operated by the Trustees of Reservations in Massachusetts has this arrangement, for instance). More of them used an arrangement which used boiler pressure -- usually in the ounces per square inch range, with provision (usually manual) for varying the setpoint to reach the desired space temperature. All of them did the job by varying the draught. All these arrangements worked with varying degrees of accuracy (it helped if the hired help remembered to stoke the boiler...). All of them were appallingly inefficient at part load, and the emissions were … pretty bad (it should be noted that at part load the radiation was never filled completely).

Then along came oil and gas burners. Then as now, these burners were either on or off for residential -- even very large residential -- applications. There was also a further demand for easy control by a thermostat which the quality could operate directly, as opposed to having the footman adjust the pressure setpoint. But... it was necessary for these controls to modulate the boiler output by on/off operation (technically pulse width and frequency modulation), and the quest was on for the same smooth even heat that the old coal burners could create.

The challenge is two fold: match the boiler output to the actual heating load of the structure, and arrange the modulation to minimize (for comfort) the heat delivered by the radiation. Fortunately, for all concerned, the radiation in the earlier days had itself a very high thermal mass (the introduction of pipe/fin tube radiation made things much more complex) and so it was -- and is -- possible to arrange for the frequency of the modulation to be rather low; many systems operate very comfortably on one cycle per hour fixed frequency, for instance. However, if one has a control arrangement which actually senses the heat loss rate of the building and, at the same time, is capable of accounting for the potential overshoot of heated, high heat capacity emitters, one can, with proper design and adjustment, achieve a simple control which varies both the frequency (essentially time between cycles) and the pulse with (firing time within a cycle) to match the demand. I honestly don't know when the first thermostats which were capable of doing that appeared; the earliest one which I've seen dated from 1930. Although remarkably accurate and precise when properly adjusted, the early ones had dry contacts and were notoriously unreliable. Along about 1945 the Honeywell Corporation developed the sealed mercury switch (actually earlier -- but for civilian use) and used that in place of the older dry contacts and the result was a thermostat -- one simple gadget -- which, when properly adjusted (I keep saying that, don't I) did exactly that with extraordinarily high reliability. In fact, a properly adjusted mercury T87, operating at a steady state setpoint, will -- with no additional control -- deliver exactly the frequency and pulse width needed to match the load.

However, a problem arises when the setpoint is varied: the adjustment is not capable of sensing that the burner on time will be too long for the demanded setpoint increase to be reached and the radiation will become full and the pressure will rise unless the boiler is exactly matched to the radiation, which is undesirable for a variety of reasons. Hence, a pressure control is also used for these conditions. This is a pressure control, not pressure safety device (the two functions should never be mixed in one control) and will take over to reduce the pulse width and often introduce a much higher frequency as well under these conditions. This is not bad; indeed, if the frequency is high enough the efficiency loss is minimal as the boiler (which also has a large thermal mass) never actually goes off steam.

The end result is a system which is fully automatic, accurate, and precise, which is controlled by two simple devices which need no operator input other than setting a desired temperature.

Unfortunately... most modern thermostats are either fixed frequency (one, two, or whatever cycles per hour) or use some form of digital simulation (often referred to as "smart control") as a surrogate for actual building heat loss rate and potential overshoot. The best of them work almost as well as the original dry contact thermostat mentioned above...

Also unfortunately, it is not common to find a system in which the time has been taken to match the boiler closely to the installed load, with the result that it produces more steam than the rest of the system can absorb. At this point engineers will -- not surprisingly -- differ as to the best method of controlling the higher frequency component of the modulation. Some, such as myself, are quite content to rely on system pressure. Others -- and I am among them for really serious mismatches -- introduce a time component. At the cost of complexity in control, this can be varied with another direct variable, such as outdoor temperature or space temperature. Alternatively, it can be essentially fixed, perhaps in two stages (a longer initial run to fill the steam mains, followed by cycling) with the timing set so that effectively, averaged over some convenient interval, the boiler output is matched to the potential load.

This also works, and the choice of control for load matching becomes a philosophical one, with no one approach being the "right" one.

Mention has been made of another problem: the mismatch of radiation to building load when the building load has been reduced. This actually should not be a problem, as it is precisely the same as operating the system under any other part load condition, as discussed above. It can be a problem, however, if the original system load balance has been altered -- such as by insulating one area but not another. This can be addressed only by changing, in some way, the radiation itself, which is another topic entirely.

mikeg2015

I suspect Coal boilers, especially vapor vacuum systems were over radiated for 1) recovery 2) capacity at low fire (partial vacuum), 3) to recover faster. Boilers were manually fed so if the fire died down early morning you needed a little extra to recover.
4) being a little over radiated meant you wouldn’t have to rely on the Primitive pressure limit controls or if it was over fired a little.

PMJ

@Jamie Hall ,

My boiler matches the installed radiation. The installed radiation is much higher than what is required to match the heat loss on design day. I know this because the system heats my house just fine at -20F with the boiler burning 30 minutes an hour and the radiation still only part full. I believe this was in fact the original design as the original boiler could not be run on "high" nor could the damper controls operate with radiators full.

My position on this is two fold:

First, that there is no practical high limit to the installed radiation as long as there is enough of it. There is no such thing as too much. Properly controlled ( at least as well as the original design) the radiation would never fill and any of it downstream of the filled part is of no consequence to anything. I think that was the point of this thread.

Second, for the same reasons there is no practical limit to boiler size once beyond the minimum needed to heat on design day. Boilers bigger than that( like mine) are easily managed with some form of PWM under which the excess size makes little difference. For anyone interested to run natural vacuum significant excess over design day capacity is an absolute requirement.

There is endless dicussion here about matching hardware and precious little about control. It should be the other way around. There are powerful and very reliable tools available the are not being utilized. While very advanced for its time the T87 won't do what is needed. I had one.

ChrisJ

From what I've heard most installations have much larger radiation than required.


Also, @Jamie Hall your boiler most certainly is not matched to the radiation if it's 33% oversized for a "piping and pickup factor" that is never used. Piping losses can be easily calculated and every person on here tells people never to do setbacks with steam so what's the point in adding in extra for pickup?

Pickup being used for more rapid recovery.....

Matched would mean the gross output matches the radiation and actual piping losses.

Jamie Hall

It's not 33% oversized. Actual measurement suggests that it is no more than 10%. More like 5. In that range anyway.

ChrisJ

> @Jamie Hall said:
> It's not 33% oversized. Actual measurement suggests that it is no more than 10%. More like 5. In that range anyway.

You and Charles installed a boiler without a 33% pickup factor? I don't recall you ever saying this before?

Jamie Hall

The boiler can be fired up to 33% -- but we've downfired it to match more closely.