"Hotwater Jack" is alive! (First data from boiler instrumentation)

In_New_England

Inspired by @fentonc 's boilertron (see here) I made Hotwater Jack in preparation for data scienceing my new modulating gas boiler to death.

Sadly, the Gods of my city have not deigned an audience with me yet to allow me the favor of pleading with them to allow a new gas connection, so I have to satisfy myself by scienceing my existing oil boiler.

This is the first data from the setup in the photo. Two sensors are attached, one to the return pipe and one to the outgoing pipe from the boiler.



It's not cold out (56F) and I forced a heat call by cranking the thermostat to 74F. The boiler cycles three times during the call, each cycle lasting about 6min.

I used the shower a bit after that but I don't see an obvious signature of that. The pipe heats up from the boiler firing to keep the indirect tank warm (is my guess) for the subsequent bumps.


A zoom into the curves during the heating call shows great similarity to the data @fentonc collected. The drop in RWT temp when the circulators start (inferred) but the water hasn't heated up yet. A rapid rise in both SWT and RWT, starting out with a 25F difference that narrows in the end to a 5F difference, presumably because the SWT is so high relative to the heat loss.

The probe is placed slightly after the section of the pipe that goes to the active zone, so we're probably measuring a little low on the SWT.

fentonc

Nice! What boiler are you using, and what kind of radiation is it driving in that graph you posted? Your boiler looks just like mine, but the "Oil" version. Also, how did you attach your temp probes to the pipes, and what sensors did you use? Also, what microcontroller did you use?

mattmia2

If it is heating an indirect and not a tankless you could set it to cold start and save some fuel, especially in the summer when the boiler doesn't need to otherwise heat and the incoming water is warmer.

In_New_England

fentonc said:

Nice! What boiler are you using, and what kind of radiation is it driving in that graph you posted? Your boiler looks just like mine, but the "Oil" version. Also, how did you attach your temp probes to the pipes, and what sensors did you use? Also, what microcontroller did you use?

It is an oil fired Weil-McLain WTGO-4 that heats a two zone forced hot water baseboard system. I believe the input is 145,000 BTU/hr.

I used zipties to attach the sensors to the pipes. I did some reading and came away with "It won't melt at 180F, but will eventually get brittle". So far so good, but the pipes, as you can see, aren't that hot for so long.

The sensors are the immersion version of the DS18B20. I got them off Amazon ("DS18B20 waterproof"). I really, really like them. They use Dallas Semiconductor's OneWire protocol. They give a digital readout of the temperature, go up to 125C and you can add many of them on the same single wire bus. The clones (which were the ones I got off Amazon), are cheap at $3.5/per.

I am using the Espressif ESP32 dev board which I also got off Amazon for $8/per ("ESP32S board"). It's the ESP32-WROOM which I think is a bit outdated. I like it better than the Arduino now - it has onboard WiFI and BlueTooth. I use the WiFi interface. Pretty fun to see the temperature readings from a TCP app on my phone.

In_New_England

mattmia2 said:

If it is heating an indirect and not a tankless you could set it to cold start and save some fuel, especially in the summer when the boiler doesn't need to otherwise heat and the incoming water is warmer.

Yes, it heats an indirect. This is a great suggestion, and I will check if I can find the control for the cold start.

I traced the piping and wiring and it looks like the following things happen

1. When the thermostat on the tank senses the water is too cold it starts up the tank circulator pump that circulates tank water through the tankless coil in the WTGO-4.
2. Sensing the temperature drop, the thermostat connected to the boiler itself tells the Taco controller to fire up the boiler, but not start any of the zone circulators.
3. There is a "Honeywell" box over the tankless coil and if I peer into it I can see the display flashing 170F, which is probably what it tries to heat the DHW to?




I'm guessing I have to fiddle inside the Taco box (that runs the zone circulators) and tell it to stop keeping the boiler warm all the time? I'm guessing the Honeywell box stuck on top of the tankless coil is the one ensuring the hot water in the DHW loop is at 170F.

mattmia2

Others know the controls better, but the aquastat on the boiler should have an upper and lower limit. turn the lower limit all the way down, that will make it cold start. That means you will also need a signal from the aquastat on the indirect to make a heat call when it turns the circulator on. if it is a 120v aquastat controlling the circulator directly you will need a relay. If it controls a 24vac control there may be an unused aux contact yo can use to fire the boiler. ideally it would be a priority zone in a zone controller or the boiler's controls but there are other ways todo it too.

fentonc

@In_New_England - I think I used six of the exact same sensors (with 9' cables!). I found that they were several degrees low (and took much longer to respond) when I just ziptied them to the copper pipe. I wound up putting a little bit of thermal grease on the interface between the sensor probe and the pipe, attaching the probe to the pipe with aluminum tape and then wrapping the taped probe with foam pipe insulation, which I then secured with a ziptie. I like the ESP32, but I find that usually the best microcontroller for the job is the one I have lying around my workshop already.

How much radiation do you have on the zone(s) that were calling for heat in that closeup graph you posted? Also, what is the high-temp limit set to, and how low does it drop before it kicks back on? My boiler is NG, but almost identical input (140K BTU/hr nominal), and the typical burns to raise the water temp 20F were somewhere between 1.5-4 minutes, depending on how many zones were calling for heat (ranging from 20K-90K of baseboard total).

pedmec

that's not an indirect. that's a storage tank

Canucker

pedmec said:

that's not an indirect. that's a storage tank

Looks like a Weil McLain Aqua plus. That's an indirect 

In_New_England

@pedmec , @Canucker it's a SuperStor 30 gallon aquabooster according to my paperwork. It's true it is not connected to the boiler as a zone, and it's not heated by circulating hot water. A pump takes the water from the tank and circulates it through the boiler's tankless coil. It looks like a janky system, but has worked for us. I'm guessing there are some energy inefficiencies involved, probably?

In_New_England

@fentonc seems you were very careful with your probes. Yeah, I just slapped them on. How did you detect that they were a few degrees low and slow to respond? Did you compare with the boiler sensors?

In that zone, I have 87' of baseboard. At 180F (which I believe the water is set to) my estimate is that that emits 50000 BTU/hr. I don't know the band setting (i.e at what temp the boiler is supposed to kick back on. In this forced run, there was a cycle every 7min.

I'm sure that there are fewer cycles during winter and imagine there is a larger difference between SWT and RWT, but my heat loss estimates are much lower than 50 MBH.

I'm looking forward to have a better matched heat source with the mod-con boiler and perhaps significantly lowered bills. We shall see.

mattmia2

Tis brings up a question with an indirect or a storage tank and a cold start high mass boiler, wouldn't you want to hold off on starting the circulator until the boiler reached a min temp, otherwise aren't you taking a significant amount of heat out of the tank to heat the boiler? I suppose stagnation helps as long as you are only involving the water near the bottom of the tank.

Dave Carpentier

Whenever I cold-started my cast iron boiler, I held off the circulators until the boiler was approaching 130f. Even though the condensation would likely be cooked off, I didnt want to push the issue.
Could probably setup a hold-off using a sensor/aquastat , like a warm-up mode ?