New Control Method

PMJ

I said I would post about the new control method I ran the second half of the season from January on. For those interested read on.

First, a reminder that it is my position that these systems originally with coal fired boilers were designed with significant extra radiation and boiler capacity. It is my contention that were it not that way, no ongoing continuous control of the output would have been possible. So even on design day, these systems will heat the house with partially full radiators. It clearly is that way in my house.

Anyway, I set about trying to run my system that same way, partially filled, experimenting with a lot of things over many years now. Partially filled radiators border on slight vacuum conditions inside them even with a big boiler firing. That is why next to no pressure at all at the header is required to keep radiators partially full. Actual pressure does not build until radiators actually start to fill completely - something that is never required to heat the house. So the project becomes how to control the boiler such that radiators are only partially full over as long periods as possible - which evens out the heat as much as possible.

I have also discussed the use of vacuum and its truly magical ability to even out the heat. I won't detail it here again. The control I ran this season used a combination of a heat sensor and a pressure(vacuum) sensor to cycle the boiler whenever the tstat was calling for heat in a way that kept rads partially full, and also adjusted the percent on vs wait of the cycles automatically without a PLC according to the conditions outside. It was easily the most comfortable heat yet, the numbers showing perhaps another 5% improvement on consumption beyond what I have already achieved. Given the reluctance of so many to give even a simple PLC a try (a shame because they really are easy to learn), this control does not require one. I did run it with my PLC doing the simple connections between these components so I could collect data, but one is not required.

This control would be installed in the control wire from the tstat to the burner and easily bypassed with one switch that could put you right back to the original vaporstat control just as it was. I installed a temperature switch at the supply pipe of my most remote radiator (in the garage) and the ultra low pressure switch at the outlet of that same radiator going into the dry return where my single vent/check valve is also installed. In the boiler room also wired with the above components is a delay off timer. Whenever the tstat is calling for heat it runs the boiler until both there is steam at the inlet to that radiator AND no vacuum PLUS 2 minutes on the adjustable delay off timer. I left it that way for 3-4 months and didn't touch a thing. On an actual cold start there is no vacuum so it runs until steam hits that far rad + 2 minutes to get a partial fill. When the system is already hot, new steam hits the radiator inlet well before vacuum is gone so the run is to no vacuum + 2 minutes. The trigger for each new burn is the inlet temperature switch dropping below 180F or so signalling that all the forward steam motion from the last burn in the vacuum has completed. So during a call for heat I just oscillated that way, starting each burn when the rads just started to actually cool and burning until all the vacuum was gone plus a small amount.

So how was this self-adjusting to the conditions? To me, this is the interesting part. First, one of the magical things about running in natural vacuum is that at the instant all the vacuum is gone each time you burn, you have just returned each and every radiator back to exactly the same fill amount it had when the burner went off the previous cycle. So, once you have a partial fill, it is very easy to maintain it with vacuum. The self adjusting part comes from the fact that the percent burn vs wait of these cycles changed a lot with the conditions all on their own. In the cold why of course the steam is condensing at a faster rate. This means that the time from when the burner last shutoff to the trigger from the temperature switch is shorter. The vacuum that got pulled in the system is slightly deeper for the same wait time because more steam is condensing per unit time so the burns must get slightly longer to kill the vacuum. When you think about it, it really doesn't take a big swing in these things to change the burn % of the total cycle time a lot. I have said all along that there is something natural and a sweet spot at around 3cph. This control operated that way all on its own - I did nothing to control the length of these cycles. I just let the steam do its thing. The temperature switch fired the boiler and the vacuum switch told me when I was partially filled back up to the same spot I was before(plus a little). So in mild weather I would get say 5-6 minute burns and 15-16 minute waits. In colder weather they would be 8-9 minute burns and 11-12 minute waits. During calls for heat this is roughly a 25% mild day burn rate each cycle and 40% in colder weather. In milder weather the time from call to satisfaction was shorter than in colder weather. Reducing the 2 minute timer would extend that time in all conditions. The longer that time is from call to satisfaction, the more even the heat is by definition. I have never required greater than 50% burn rate even at -20F to heat the place with my big boiler.

I will note here that the 2 minutes additional time is arbitrary and easily adjustable. It is needed only to overcome leaks and push that small amount of air back out each time so you don't lose ground. The longer you make that time, the more your system becomes just like a pressure one - if you make it long enough you will fill your rads and pressure controls will take over. I just chose that number and never touched it for 3 months. I think one minute would have done fine. It is not critical. It just has to be enough that over enough cycles the system will gradually fill as opposed to emptying out due to air leaks. The leak rate is not related to the conditions. Need to recover from a power outage or something? No problem, throw the bypass switch.

So to sum up this control requires 4 components on a 2 pipe system: the temperature switch, the pressure switch, the check valve, and the delay off timer. These things don't total $200. It is pretty simple and easily bypassed. It is worth the effort? I already know the opinion of many here is no. All I can say for those interested is that the comfort level is dramatically better than my vented vaporstat system and overall efficiency improved 30% or so from that same system.

Jamie Hall

Interesting and worthwhile concept. Probably the biggest benefit comes from off-design days -- and perhaps in your situation the actual design outside temperature -- that is full capacity -- is somewhat or even considerably below the real lowest temperature/maximum demand your system is capable of (not true for me!), in which case the benefit would be even greater.

I particularly like the fact that it doesn't depend on anything really fancy, and the operation is easily described. This is almost essential in my view, as you know, as one can't guarantee that the person doing the install will always be the person figuring it out at oh dark hundred some winter morning!

PMJ

> @Jamie Hall said:
> Interesting and worthwhile concept. Probably the biggest benefit comes from off-design days -- and perhaps in your situation the actual design outside temperature -- that is full capacity -- is somewhat or even considerably below the real lowest temperature/maximum demand your system is capable of (not true for me!), in which case the benefit would be even greater.
>
> I particularly like the fact that it doesn't depend on anything really fancy, and the operation is easily described. This is almost essential in my view, as you know, as one can't guarantee that the person doing the install will always be the person figuring it out at oh dark hundred some winter morning!

That's just it Jamie, all days are off design days. With everything I consider I find the system far easier to run with extra boiler on hand. Small changes in boiler firing rate then are bigger % changes in output relative to the demand and thus easier to achieve. I am convinced it was not possible to run the coal fired boiler anywhere near its true capacity with the largest fire one could possibly make. Rather a bigger boiler was run comfortably nearer the middle to lower end range of its capacity - a manageable project.

And I agree with you that simplicity is an advantage. A little documentation goes a long way. This one should have the bypass switch which turns the whole thing into the piece of wire going from the tstat to the controls everyone understands - just like it was before.

Motorapido

How could this be applied to one-pipe steam?

MarkS

@PMJ if I understand you correctly, this is a variation on your original "timed pulse" control strategy, this time responding to environmental inputs. Nice work!

With the PLC and an indoor temperature sensor, you could determine the difference between the measured temp and the setpoint, then automatically run in bypass mode if conditions warrant.

Have you timed from burner off to when the temperature switch opens? Is it fairly consistent?

PMJ

MarkS said:

@PMJ if I understand you correctly, this is a variation on your original "timed pulse" control strategy, this time responding to environmental inputs. Nice work!

With the PLC and an indoor temperature sensor, you could determine the difference between the measured temp and the setpoint, then automatically run in bypass mode if conditions warrant.

Have you timed from burner off to when the temperature switch opens? Is it fairly consistent?

Mark,

Yes, this is a variation on my long time pulse strategy. Given the large resistance here to anything fancy I just started thinking about doing what I was doing without a PLC and try to add something that would respond automatically to the demand, keeping radiators partly full and warm over long periods (the definition of even heat). I have been really happy with what I've found.

The time from burner off to temp switch open ( I call it the wait period) is quite consistent during a single call for heat. But this is what changes over time as demand rises or falls. In the cold when steam is condensing faster it takes less time from when the fire goes out to consume what the system has available and this time gets shorter. In mild weather it is longer. The burn times change too then the opposite way. The interesting thing is that the total cycle time remains in a range of 18-22 minutes or about 3CPH no matter what. Over years I ended up forcing a 20 minute cycle because it just "felt" like it was the most efficient. That the system oscillating on its own chose the same cycle seemed to me confirmation that there was something special about it.

I will also say here again about the vacuum being so effective at controlling the fill level of the system. It is almost startling how on each new burn, at exactly the moment all the vacuum from the previous cycle is gone, each and every radiator is filled exactly to where it was the last time the burner shut off. It might be the most significant thing about this control method. Over time rads in the colder areas (upstairs vs downstairs etc) gradually fill more or less relative to each other, but with each new burn you get back very close to the previous one everywhere. And the exact total amount of vacuum is not important at all. Whatever vacuum a leaky system created in each wait period, when that vacuum is gone on the next burn the fill level everywhere is the same. Really cool (to me anyway).

PMJ

Motorapido said:

How could this be applied to one-pipe steam?

TImed 3CPH pulsing can be applied to one pipe systems any time and I highly recommend that at a minimum.

It is the vacuum part that is harder in one pipe and much discussion about it goes on here from time to time. All rads need a check valve somehow or all rad vent ports run in air line to a common location. I will simply tell you that if I had a one pipe system I would be trying something. The advantages to vacuum to to great to ignore.

Steamhead

PMJ said:

It is worth the effort? I already know the opinion of many here is no.

Not necessarily, @PMJ .

But I've said this before, to @IgorZhadanovsky and others, that as contractors we need a clearly defined package, with a clearly defined installation method, at a clearly defined price for this to sell. Customers don't seem to want anything that requires experimentation- they want a one-and-done scenario. We even run into this sort of resistance when discussing outdoor-reset controls on hot-water systems.

I really admire the work you all have been doing. The fact that you've simplified things is a positive step. 30% greater efficiency is a worthwhile target. Now it needs to be packaged into something we can sell.

PMJ

Steamhead said:

PMJ said:

It is worth the effort? I already know the opinion of many here is no.

Not necessarily, @PMJ .

But I've said this before, to @IgorZhadanovsky and others, that as contractors we need a clearly defined package, with a clearly defined installation method, at a clearly defined price for this to sell. Customers don't seem to want anything that requires experimentation- they want a one-and-done scenario. We even run into this sort of resistance when discussing outdoor-reset controls on hot-water systems.

I really admire the work you all have been doing. The fact that you've simplified things is a positive step. 30% greater efficiency is a worthwhile target. Now it needs to be packaged into something we can sell.

I do hear you @Steamhead , and I get that contractors really need to keep moving - and that few customers will pay for much extra "adjusting".

I also get that I have the huge advantage of being a live in steam man. I have an enormous number of hours in observation that customers surely won't pay a contractor for - they probably even find it upsetting if you want to go all over the place in every room multiple times to see what is happening. My wife thinks I'm nuts by the way as I am always wandering around to check things (she does like the heat though).

So I hope all the contractors out there don't read me as being critical or blaming them for the lack of development. I do get the spot you are in. I don't think the product is coming for you to use - I think the die is cast on this. There are so many systems out there it will take a while to replace them all. But new investment in developing new controls for them on a commercial level is unlikely to happen in the face of no new systems going in.

So mostly I am just speaking to folks who want to mess with their own systems. There is a lot you can do fairly easily and get a lot of improvement - especially in 2 pipe.

I am sorry none of this vacuum stuff survived the switch to intermittent fire. I would put my all in annual efficiency numbers(fuel and total equipment cost) up against and kind of heating system.

Steamhead

Only idiots tear out steam systems, @PMJ . What would it take to get this unit contractor-ready?

Motorapido

PMJ said:

Motorapido said:

How could this be applied to one-pipe steam?

TImed 3CPH pulsing can be applied to one pipe systems any time and I highly recommend that at a minimum.

PMJ -- I'm an excellent mechanic but not all that swift with electronics. So I need a little consultation. For my one-pipe system, can you suggest a particular duty-cycle timer and suggest the best method for wiring it into my control system? For instance, wired between stat signal wire to boiler, or wired after the vaporstat, to delay re-start when the vaporstat first cycles on pressure (I'm about 55% oversized, extremely well vented on mains, totally insulated everywhere, etc., so my only remaining optimization is a timer addition. Everybody tells me retrofitting my peerless to a two-stage burner is problematic and might not yield adequate return on investment, so I'd like to add a timer).

PMJ

Motorapido said:

PMJ said:

Motorapido said:

How could this be applied to one-pipe steam?

TImed 3CPH pulsing can be applied to one pipe systems any time and I highly recommend that at a minimum.

PMJ -- I'm an excellent mechanic but not all that swift with electronics. So I need a little consultation. For my one-pipe system, can you suggest a particular duty-cycle timer and suggest the best method for wiring it into my control system? For instance, wired between stat signal wire to boiler, or wired after the vaporstat, to delay re-start when the vaporstat first cycles on pressure (I'm about 55% oversized, extremely well vented on mains, totally insulated everywhere, etc., so my only remaining optimization is a timer addition. Everybody tells me retrofitting my peerless to a two-stage burner is problematic and might not yield adequate return on investment, so I'd like to add a timer).

@Motorapido ,

Study the attached PDF. These are available at Grainger for about $66. + $10 mounting base.

You wire this on the 24vac signal line from the Tstat before it gets to the boiler. I or several others who have done this can help with a wiring diagram. Conceptually this little box just changes the continuous signal from the tsat to an intermittent one, with an on/off schedule adjustable by you. With your significantly oversize boiler I suggest you set it to an on/off cycle that gives you an overall on percentage run time you know will heat on design day - like 50% and a CPH of 3 so 10 minutes on/ 10 minutes off. It will be a little slow in recovery mode, but way better when the system is already hot. I ran with one of these for several years during the start of my experimentation. At a minimum, you and everyone else running a steam system knows the feeling of knowing your rads have already filled enough to meet the current demand, but then wishing your tstat knew it too. This little timer is the simplest way to give a little wait and allow your tstat to get the right idea and reduce your overshoot.

You also wire in a bypass switch to skip the whole thing and easily go back to the way you are.

PMJ

Steamhead said:

Only idiots tear out steam systems, @PMJ . What would it take to get this unit contractor-ready?

@Steamhead ,

If "contractor ready" means an off the shelf commercially available product as applies to everything you sell now then the answer is that it isn't happening. I just don't see anyone or any company producing and supporting such a device for residential steam the conventional way you expect and do so for good reasons.

Given what I view as the precarious perch residential steam heat is on these days, the use of such a control by a contractor would require, I believe, a situation where that contractor simply purchased the few simple components on his own and installed them. From the technical side it is not complicated at all in two pipe and easily within the skill set of craftsmen like you. Legally, I don't know what liabilities there are. As a homeowner I have been assured that my insurance policy has no "stupidity clause", as my agent of many decades puts it, and that my own work on my own system does not in any way put my coverage at risk. As a hired professional you would know better than I how that works and if doing what I propose would be a show stopper.

Be all that as it may, I do wish someone like you would spend the $100 bucks or so and try your hand at a PLC. I promise you it is easier than you think and a truly amazing bang for the buck. Reliability in a whole different league than a PC running Windows, and so inexpensive that in the rare case of failure you would be happy to have pre-programed spares on your truck always. At my plant I have several running that have never been restarted/rebooted for more than 2 decades.

What I did not mention in my response to @Motorapido but will here is that for almost exactly the same price as the unit I suggested there, you can have a PLC that has every timer on that page in it already. Beyond that, all contacts that are physically normally open or closed in the field can become either way inside the PLC at your convenience as you wish with only keystrokes. Those of us who use them every day would gladly have one instead of any "simple" circuit of the wires and relays that would be required to execute the control I have proposed here. It would be considerably less expensive done with a PLC and considerably more reliable. I truly believe if one of you would take the leap on these a whole new world of opportunity would be opened to you and you would become a hero to the first customer who would allow you a little space to work. Frankly speaking I don't see any other way forward on this.

I am always happy to share whatever I can and will attempt to draw up a hard wired version of what I propose if need be. Others such as MarkS are far more skilled in the actual execution of designing and building such devices than I am.

MarkS

PMJ said:

Be all that as it may, I do wish someone like you would spend the $100 bucks or so and try your hand at a PLC. I promise you it is easier than you think and a truly amazing bang for the buck.

If anyone wants to try their hand at this, I have surplus inventory from the EcoSteam project available for sale. PLCs, power supplies, terminal blocks, indoor and outdoor temperature sensors, and several fully assembled & tested systems. PM me for details.