why a PID thermostat is needed

R. Kalia

I thought I would start a new thread to explain why an intelligent, analog-output thermostat is needed with new boiler designs.

In a basic system one has an aquastat to set water temperature; to make it more complicated, one puts in outdoor reset to vary the water temperature. So one needs to figure out a reset curve that relates the water temperature to the outdoor temperature and desired indoor temperature. But outdoor reset is an unnecessarily indirect and complex solution, resulting from an obsession with water temperature. The real goal is to control the room temperature; one should not have to think about the water temperature at all.

Boilerpro has already described, deep in another recent thread ('enhancing baseboard output'), how one directly controls the room temperature. I will describe it in my own words. All you do is set the circulator to run constantly, then use the room thermostat to switch the boiler on and off. You don't have to think about the water temperature, because under the room thermostat's control it will automatically reach and hold whatever temperature is needed to keep the room at 70. The water won't get any hotter than what is needed, which is the goal of outdoor reset as well, but here it happens automatically i.e. without having to guess the reset curve. If the sun shines in, the water temperature will drop (try getting that with outdoor reset). You don't even need P/S piping because there is always full flow.

With me so far? Next step, modulating boilers. These don't just go on and off; they can vary their output. So in the scheme just described, the boiler should be controlled by an analog signal (e.g. 0-10vdc, 0-20mA, etc) that runs them up and down rather than on and off, increasing efficiency by eliminating cycling. Hence, PID thermostats.

Now admittedly the current generation of modulating boilers don't allow direct access to the modulation; instead, they have internal logic to set the modulation themselves based on how far away they are from the aquastat setting (again, it's that obsession with water temperature). In the scheme just described, the aquastat setting is irrelevant except as a safety limit. The water temperature will never normally reach the boiler's aquastat setting. Modulating boiler manufacturers will need to make the modulation externally controllable. Meanwhile, Honeywell will need to make thermostats that aren't just on/off.

bill clinton_3

some doubts

The scheme of letting the thermostat fire the boiler and expecting the boiler temperature will settle into just the right water temperature to supply the heat needed is an old one. The Hydronic Institute started promoting it in the fifties, recommending two stage thermostats: first stage starts pump, second stage fires boiler.

I took an IBR class in the sevienties and immediately tried the strategy. It didn't work. The boiler rode the high limit at all times. The reason was that the boiler produced heat much faster than the system could disperse it. This will be the case most of the year and will be especially a problem with multiple zone systems. For it to work, you need high mass systems with little zoning and heat emitters capable of meeting or preferably exceeding boiler output.

Bill

R. Kalia

going hunting

> The boiler rode the high limit at all

> times. The reason was that the boiler produced

> heat much faster than the system could disperse

> it. This will be the case most of the year...



This problem is well-known in the process control field where it is called 'hunting'. It is inevitable when there is 'dumb' on/off control of a 1970's (oversized) boiler.

Of course things will work perfectly with a modulating boiler, but that is not essential---just using a thermostat with sufficient anticipation, or a digital thermostat with cycle-rate control, will result in much better behavior. PID is even better but the cure for the problem you mentioned is anticipation (just the D in PID).

Thermostats are direct and simple, as opposed to outdoor reset, which is (to a temperature control person) a very roundabout and imperfect way to approach the problem.

Me_8

Just how would more than one zone or even using an indirect be possible ? 0-10vdc and 4-20 ma stats do exist can't remember the name but have seen them controlling 3 way valves.

R. Kalia

zones

> Just how would more than one zone or even using

> an indirect be possible ? 0-10vdc and 4-20 ma

> stats do exist can't remember the name but have

> seen them controlling 3 way valves.


It works with zones only via mixing valves---outdoor reset is the same way, i.e. if you want continuous circulation, and who doesn't, you can't do zones unless you do different water temperatures for each zone.

Dave_22

Some thoughts...

"All you do is set the circulator to run constantly, then use the room thermostat to switch the boiler on and off"
*** I thought for better efficiencies, you want longer run times and NOT to cycle boiler on and off. With a modulating boiler, this is much easier whether you have outdoor reset or not. Reset helps with a modulator by not heating the zone up so fast and getting a longer run time.

"Next step, modulating boilers. These don't just go on and off; they can vary their output. So in the scheme just described, the boiler should be controlled by an analog signal (e.g. 0-10vdc, 0-20mA, etc) that runs them up and down rather than on and off, increasing efficiency by eliminating cycling. Hence, PID thermostats."
*** Yes, i agree. But return temp, i would think, is proportional to room temp. The hotter the room gets, the smaller the delta T of the return water, and the boiler modulates to low. This to me, is the same thing a PID thermostat would do. The hotter the room gets, the lower the modulation of the boiler would be.

"If the sun shines in, the water temperature will drop (try getting that with outdoor reset)"
*** Yes, and again, the return water temp would rise which would lower the modulation based on the return temp.

I'm just thinking out loud here, but I think since the return temp is kind of proportional to the room air temp, it almost IS like having a room PID thermostat. And this is because the larger the delta T between water and room temp, the more heat that is pulled from the water. The smaller the delta T between water and room temp, the higher the return water temp. I also think that outdoor reset on a modulator really helps to lengthen your run times and stay closer to setpoint without overshooting. OK- sorry!!!! Just some thoughts going through my head!!!!!

bob_25

Oh contraire

come on over sometime, it works great. I have a slant-fin gas cast iron boiler and slant-fin base board sized for 150F water. I was going to hydro pulse school at the time and sized the baseboard for a pulse. Fortunatily I couldn't afford the pulse:) Water temp modulates from 100 to 160. Contrary to what everyone says it doesn't condense. The pump has been runing since 1984, I'm to lazy to shut it off in the summer. For zoneing I have a master zone for main living area and baths that is continious, this stat controls the burner only. Bedrooms have zone valves but have to make do with whatever water temp is available. Cheap, smooth and efficient. bob

jw

HTP's VisionII

Does exactly this! Modulates mixing on each zone and integrates it into the munchkins modulation and outdoor reset. (would love to see THAT progam!)

R. Kalia

return temp

> "All you do is set the circulator to run

> constantly, then use the room thermostat to

> switch the boiler on and off" *** I thought for

> better efficiencies, you want longer run times

> and NOT to cycle boiler on and off. With a

> modulating boiler, this is much easier whether

> you have outdoor reset or not. Reset helps with a

> modulator by not heating the zone up so fast and

> getting a longer run time.


This setup is the same as reset in that regard, because the water temperature will only get as hot as is needed to get the room hot...in other words the water will be as cool as possible.

> "Next step,

> modulating boilers. These don't just go on and

> off; they can vary their output. So in the scheme

> just described, the boiler should be controlled

> by an analog signal (e.g. 0-10vdc, 0-20mA, etc)

> that runs them up and down rather than on and

> off, increasing efficiency by eliminating

> cycling. Hence, PID thermostats." *** Yes, i

> agree. But return temp, i would think, is

> proportional to room temp. The hotter the room

> gets, the smaller the delta T of the return

> water, and the boiler modulates to low. This to

> me, is the same thing a PID thermostat would do.


No, it isn't. The return temperature depends somewhat on room temperature but is not proportional to it. And modulation isn't set up to be really sensitive to the room temperature, the way it would be if controlled by a PID thermostat. You can't control based on return temp, too many unpredictable variables (it is different for each system, and even in a given system, if someone turns a radiator off, that affects the return temperature).

The only measurement of any value is the room temperature---the final product---hence a thermostat is the right control device.

ALH_3

Centralization

Outdoor temperature is not irrelevant, and water temperature is relevant. You need two temperatures to calculate heat transfer rate. (obsession with water temp?)Primary/Secondary plumbing is more about protecting the boiler from cold return temps than flow restriction in a cast iron boiler isn't it? Modulating boilers like the Vitodens are very likely to require primary/secondary plumbing because of the flow restriction of the heat exchanger.

It seems to me that due to expense, it is a lot cheaper to have a control in the mechanical room that integrates several room temperature sensors, outdoor temperature, system temperature, boiler temperature, and DHW demand. Obviously the heat load can be very different in different rooms at the same time. Would the thermostats also have contacts to open/close zone valves as high limits for each zone?

Keeping the electronics as centralized as possible with thermisters in each room seems the most straigtforward approach to me.

PJO_5

\"Who doesn't?\"...

This whole idea sounds great, and is based on using constant circulation. I don't feel that it's absolutely needed...I (nearly) use it on design days, then slowly pull away from it the rest of the season. I'd guess the average run time of a typical circ in my home is about eight hours/day over the heating season. My radiant zones a bit more(one circ w/ three zone valves), my baseboard zones a lot less (three circs), and my primary pump the most.

The reason? Five circs running 24/7/200 days (or more) seems ludicrous to me. I've seen systems on this great Wall that have twenty circulators...while most people may not think of it this way, the TOTAL system efficiency should include the electrical usage IMHO. The more zones in a home, the more electrical usage you have running circs. Now, if you have PV panels that's a different story...

I don't give up any comfort doing it this way, but there is an occasional set of noises from the upstairs baseboard system as she "settles in" for the evening :-) The passive solar in my house is very good, and that comes into play when I rest the programming of my t-stats and the boiler temperature.

I'm not a contractor, and I think it's a very good idea in many ways. Just not for me...

Take Care, PJO

hr

But an outdoor reset

brings so much more to the table.

You get near constant circulation if they are designed and set up correctly.

You get return temperature monitoring and protection.

You get anticipation.

You get DHW priority.

You get lots "adjustability"

Many have some degree of data logging, nice digital displays.

The cost is very reasonable. Nice reset controls can be had for 170 bucks or so.

Why not? I'm sure the method you describes will have some of these features, but it is still a bang/ bang control logic.

hot rod

To Learn More About This Professional, Click Here to Visit Their Ad in "Find A Professional"

R. Kalia

> Outdoor temperature is not irrelevant, and water

> temperature is relevant. You need two

> temperatures to calculate heat transfer rate.


Why do we need to calculate heat transfer rate? in the scheme I described, the thermostat tells you this so you don't have to calculate it, and the system automatically finds the correct reset curve water temperature.

> (obsession with water temp?)Primary/Secondary

> plumbing is more about protecting the boiler from

> cold return temps than flow restriction in a cast

> iron boiler isn't it?


Cast iron boilers cannot use the scheme I described because the water won't always get hot enough. It is for modern boilers only.

>

> Obviously the heat

> load can be very different in different rooms at

> the same time. Would the thermostats also have

> contacts to open/close zone valves as high limits

> for each zone?


I don't know of an outdoor reset system that does this either. Individual room control is best achieved with TRVs, whatever the central control system.

ALH_3

v&t

It sounds as if a combination of Vitodens and tekmar 370 would accomplish what you're describing. It's not really possible, because the tekmar couldnt talk to the boiler to control the system temp. If Viessmann and tekmar could just work together I think it's all possible.

Duncan_15

Outdoor reset and PID questions.

I like the idea. Like Bill, I also saw this strategy offered by I=B=R, and it makes sense. Are you sure it wasn't meant for cast iron boilers? I saw an old (60's?) paper on it. It showed fuel savings over on/off cold start controls in the 20% range.

And like Bill pointed out, there can be problems with it.

OK. PID means Proportional, Integral, Derivative. They all use a setpoint - the thermostat setting or target temperature. And a measurement - actual room temperature.

As I understand it, what makes the PID control more accurate is how it handles *the difference between the setpoint and measurement*, called the error.

Proportional: The output of the controller is proportional to how big the error is.

Integral: The output of the controller is proportional to the amount of time the error has been present.

Derivative: The output of the controller is proportional to how quickly the error is changing.

I'm thinking our scenario with on/off controls rather than gradually changing (analog or modulating) controls are simply a crude approximation of PID control. The water temperature of the system would correspond to the "Proportional" part of PID, also known as the gain. If the gain is too great, the control loop goes unstable.

The integral part would be the limit, which adjusts for load change, and the derivative part, I think you're saying, is the thermostat anticipator.

Wouldn't boilers that modulate around the setpoint just be dampening the gain? Maybe a little too late? A slight refinement over letting the limit control the gain?

And wouldn't outdoor reset's role be to lessen the Proportional part? That is, instead of a huge gain (high water temperature) which makes the control loop unstable, wouldn't the lower temperature of the outdoor reset make the gain more controllable and eliminate the problem Bill ran into with his boiler hitting the limit all the time?

That is, wouldn't the temperature be smoother and more even with less room temperature excursions? Seems to me, this would be a fair advantage over running up to the setpoint.

From everyone I know who lives with outdoor reset and constant (or near constant) circulation, this is the case.

Tekmar's outdoor reset with indoor room temperature units (analog RTUs), does indeed have PID logic, if I recall correctly. Unfortunately, it still controls the water temperature with an on/off gas valve. But it looks at the size of the error, how fast it's changing, and how long it's been changing.

Someday, there will be a PID interface for a modulating boiler for residential.