Timer with OD temp compensation timing

tim smith

Talking to a control company who can make an electronic time clock with od temp compensation for on times, either interval length or occupied time of day. Say for every 5 degrees below base design temp it would lengthen time say 10 or 15 minutes. Wondering if anyone see's a need for this for light commercial work, predominantly mid size multi family. We have quite a few buildings who just run on time clocks as that is what ownership wants. These are mostly steam systems and or old hot water boiler systems. I know we could use a heat timer or Tekmar but this is a much more simplified system. Any thoughts?

EdTheHeaterMan

This feature is built into just about every Honeywell setback thermostat for the last 20 years.  Honeywell calls it “adaptive intelligent recovery”

but if you want to reinvent the wheel ... go at it 

mattmia2

Does the honeywell actually use the outdoor temp? I remember when I bought my honeywell about 20 years ago and went through the trouble of getting the outdoor thermistor and hooking it up, I later found out from a honeywell rep here that it just remembers how fast it responded over the past couple days.

Big Ed_4

look into Heat Timer

ratio

It uses it indirectly, the lengths of the previous burns are directly related to the outdoor temp. IIRC the only time a Honeywell stat (or at least, the one I was looking at) uses the outdoor temp was for changeover on a dual-fuel setup.

@tim smith, I think you're looking for a PLC. I don't claim to know much about anything, but I can't say as I've ever come across a timer that changed its interval based on temperature. Honeywell does make a flavor of the T775 with reset, it has both an internal schedule and a digital setback input. Since it actually does have a timeclock in it, I might try calling it a ' temperature responsive programmable timing relay with unoccupied setback scheduling capabilities', or some such. YMMV.

Jamie Hall

See if @PMJ has some thoughts on this one. He may even have some sources. I know he's done a good bit of good work with timers on steam systems.

If done properly, the idea might work well for multi-occupancy buildings. I myself would be a bit concerned about the system not being able to compensate for other influences on heat loss -- particularly wind and solar losses and gains -- so it might be necessary to have some sort of thermostat control as well. Or perhaps TRVs for individual spaces.

Just thinking out loud here. Dangerous habit...

EdTheHeaterMan

mattmia2 said:

Does the honeywell actually use the outdoor temp? I remember when I bought my honeywell about 20 years ago and went through the trouble of getting the outdoor thermistor and hooking it up, I later found out from a honeywell rep here that it just remembers how fast it responded over the past couple days.

No outdoor sensor. there is an algorithm they use to predict the recovery time. if the time is exceeded then the recovery starts sooner. If it takes less time, the time is reduced. It remembers the previous three days and calculates accordingly.

I'm wondering what the difference is between reading the outdoor temperature and using the timer method in the thermostat is... if the results are the same?

If your decision is based on your feelings of the technical aspects as opposed to facts, then I can't use the facts to convince you that the results are similar with both technologies. You should just go with your gut feeling on this issue.

tim smith

Thanks for comments all. The main reason for this is there is never a good apt in these buildings and or hallway to sense temps. The building owners on these 15 to 50 unit buildings won't go for Thermostatic valves. They are old boilers, just trying to make sense of them as they stand. Usually the straight time clock method works ok in the shoulder seasons but when we get the intermittent cold, they are always scrambling. I have just always felt with using the tech (albeit limited) that already is being used, the introduction of a little programming and addl OD sensor could prove useful. In an ideal world I would like 4 ons and 4 offs. Then use the OD feedback to adjust the on time I think as temp drops. I have used heat timer and Tekmar, just feel there could be a simple way to accomplish this. Of course this is not needed with our outdoor reset on current modcons. Solar gain can be a kicker in any scheme. The controls company I talked to are open to the idea if there is a bit of demand. Just tossing some bait out to see if any interest.

ratio

The Honeywell 8000 series can use up to IIRC 5 wireless remote indoor temp sensors. If that's not enough sense points, anything that can use a thermistor can be set up for averaging a square number of sensors (4, 9, 16, etc.) trivially, and an oddball number with some slight gyrations. The same thing can be done with the T775, albeit with Pt100 sensors.

mattmia2

In michigan it is frequently 50 one day and 20 the next. i don't see how measuring burn time is going to measure that.

PMJ

Nice to see some thinking going on about this.

The best approach I have seen to this on paper was Ecosteam. PLC based using some basic inputs about the heat loss based on outside temperature it calculated a run time since the boiler output was known to provide an appropriate run length for the current conditions. I arrived too late to try one. This market had already expressed its disinterest, as it seems to for any changes to the status quo.

I have resisted going to an outdoor sensor in an effort to stay simple. The original systems continuously matched steam supply to heat loss in real time simply reacting to the changes in the ultra low pressure in the system. It turns out that the rate of steam condensation in radiator through the same filled area varies a lot with the conditions. PLC based myself, I have used timers to maintain a more steady fill amount, which necessarily requires matching supply more closely to demand.

The fact we all know is that the burn time required to get steam to the radiators depends more on how long it has been since the boiler last ran than it does the conditions. It is for this reason that I have ended up running every burn until steam actually reaches the most remote point I need it to. I time a set fill burn time from there. This gets me to a known partial fill amount no matter the conditions or the relative coldness of the start. This burn time varies on its own then, because it is terminated by a remote temperature sensor plus fixed timer.

The remote sensor also initiates the next burn when it opens up if the call is still in progress, and runs back to the same fill level. The time between the burns then is determined by the conditions. As I said, steam condenses faster through the same filled area in colder conditions and so time between burns gets shorter. Cycles vary on their own matching boiler hourly output more closely to what is needed. This control reacts to conditions in the system - more along the lines of the original design with no other outside input.

There are many ways to skin this cat. @tim smith , great to see some thought going in to this.

Chris_L

As @PMJ says, needed run time for a steam system is largely dependent on how long it has been off (i.e., how cold the system is).

I have taken the somewhat different approach of turning my two boilers off at a low pressure (7-10 inches of water) and keeping them off for a fixed period, about 15 minutes, before allowing the thermostat to control again.

I might be interested in the device @tim smith mentioned if it could also be used to vary the off time, to reduce the amount of overshooting of the thermostat setpoint in warmer weather.

I have to wonder, though, in these days of "smart" thermostats, don't any have virtual outdoor sensors to pull the local temperature or even wind an solar insolation data from the internet?

Since the microprocessors I use are connected to the internet, I may play around with this to use the data to vary their off times.