Simulating my home heating system for fun and profit

fentonc

In my on-going quest to understand my home hydronic heating setup, I've been gradually increasing the capabilities of the controls I've added. First I was just tweaking the thermostats and aquastat settings. Last season I added a PLC to do zone synchronization and add a limited thermal purge mechanism, and this season I wired the PLC to be able to inhibit the burner as-needed in order to have more control over the purging. A common theme on this forum is that it's hard to tell if and how much some change to a system might actually help (or hurt!) things. With my more fine-grained boiler controls, I was running some thermal purging experiments when I realized that the monitoring system I built was pretty close to being able to just simulate my whole heating system. For the last two winters, I've recorded 1) hourly outdoor temperature data, 2) gas usage and 3) measured heat delivered to each zone throughout the day. Using that data, I made a physics-based simulator of my setup:

• a cold-start boiler with an internal aquastat, fixed firing rate and calibrated heat-capacity and standby losses
• three zones, each of which have their own baseboards (of different types/lengths) to output heat, internal water volume to store BTUs, virtual thermostats with hysteresis, and zone valves to move BTUs between the boiler and the zone.
• A separate 'controller' where my PLC sits, which can monitor the thermostats, inhibit the burner and control the zone valves as needed

By 'replaying' the external temperature and per-zone heat-loss data through my simulator, I can simulate the last 2 winters second-by-second and do virtual experiments to check the impact on gas usage, efficiency, over/undershoot of temperature, etc. Thinking about changing your supply water temp? Right-sizing your boiler? Adding buffer tanks or more baseboard? Fancy control schemes? Wonder no more! If I simulate the exact control scheme I used for the previous winters, I can match my actual gas usage to within ~2%, which is about what my utility guarantees for the energy content of the gas. Given that it's simulating a whole winter 1 second at a time, it's far more accurate than I expected when I started working on this.

So what to do with an extremely accurate winter simulator? Let's play fantasy plumber!

First, let's look at the impact of changing the aquastat high-limit (with fixed 20F differential):

My boiler aquastat had been set to 180F when it was installed, but in 2022-2023, I lowered it to 160F. So lowering my aquastat high-limit by 20F lowered my gas usage by about 5.7%, and lowering it further to 140F would have brought the savings to nearly 13% - neat!

Last season I got more serious about mucking with the controls, and I added zone-synchronization as well as a limited thermal purge down to about 140F (the aquastat low-limit) at the end of each cycle, but left the aquastat at 160F. If I replay the 2023-2024 winter under different scenarios, I can see my extra controls managed save about 18% vs the 180F default settings, and about 12.8% vs the settings from the previous winter.

By doing a second-by-second simulation, it let's me capture how dynamic the effects of different control schemes are, and what an impact it can have on total fuel usage. I've still got many more experiments to try for the upcoming heating season!

Big Ed_4

What is a PLC ?

hot_rod

https://forum.heatinghelp.com/discussion/comment/1811649#Comment_1811649

programmable logic control

fentonc

@Big Ed_4 - It's a Programmable Logic Controller, which in this case means a little computer that can monitor 8 inputs and control 4 relays as outputs. It sits between my thermostats and my boiler, and the output relays control the 3 zone valves and the auxiliary limit circuit on the boiler (wired in series with the rollout switch and the flue damper).

Here is an example of the water temperatures of all three zones actually measured during a cold-start burn vs the equivalent in my simulator (although actual room temperatures and heat loss don't match up exaclty):

vs

Big Ed_4

I am thinking you are running the system on a low mass boiler ?

fentonc

@Big Ed_4 - It's a Weil-McClain CGa-5-PiDN cast iron boiler. Nominally 140k BTU/hr input, and 2.7 gallons of water inside it, connected to 3 zones of fin-tube baseboard. Here's a big overview from the first season when I added my monitoring system:

https://forum.heatinghelp.com/discussion/188257/end-of-season-report-hows-my-boiler-doing

Big Ed_4

Thanks

fentonc

Now let's take a closer look at thermal purging - this is where we continue to run the circulator after the last thermostat has been satisfied, in an attempt to move as many BTUs as possible out of the boiler and into the conditioned space. This time I disabled the zone synchronization, left the aquastat high limit at 160F with the 20F differential, and I swept the 'low limit' for purging (i.e. I ran the circulator until the boiler thermostat hit the low limit).

In reality, I used a 140F purge limit but I had zone synchronization enabled, so my simulated gas usage was about 165 therms (and 169 when I actually measured it), rather than the 183 therms with a 140F purge limit but no zone synchronization. Purging down to 100F would have achieved the same result as the zone synchronization (~165 therms), but with a less complicated control scheme. Purging all the way down to room temperature could have shaved another ~6% off of my gas usage, knocking it down to 155 therms.

Thermal purging works pretty well!

EdTheHeaterMan

This is interesting stuff…. The folks at Brookhaven labs can take the winter off now!

fentonc

But we can do even better! Our losses derive from the degree-seconds that the boiler spends at elevated temperatures - if we can purge things quicker, we can benefit even more. Now rather than purging into the last zone to call for heat, what if we ran our purge cycle with all three zones open? This has a number of benefits - it brings more feet of radiator to bear when it comes to dissipating the heat, it lets us take advantage of the colder water sitting in the inactive radiators, and it spread the heat out more gently, so there is less risk of overheating.

And, best of all, when we purge it all the way down to room temperature, it shaves another 10 therms (6%) of gas off of our previous-best record. This is a full 27.8% lower gas usage than the 'default' settings and control scheme we started with - not bad for just some fiddling with controls!

Mad Dog_2

Great little case study...Mad Dog

DCContrarian

Not to be Contrarian, but the only reason this helps is that your boiler is in unconditioned space. If you insulated your basement — as modern codes require — none of this would matter.

fentonc

That's the kind of armchair advice we can test now =) How much would we need to reduce aggregate boiler heat loss, given the default 180F aquastat settings, to match the gas usage of my best proposed control scheme thus far?

So to match the ~145 therms of gas usage from my 'all-zone purging' scheme while leaving the controls untouched, we would need to reduce the aggregate boiler standy heat loss by 80%. I don't know where exactly all those BTUs are going, but that doesn't sound like it would be straightforward.

DCContrarian

When the basement is insulated, those BTU's go through the basement ceiling and into the conditioned space.

fentonc

Well, some of them would (and the more I insulate and air-seal, more of them would). Some of them already do, of course, and are baked into the parameters I fit. But I still have a flue, and it's still an atmospheric boiler that needs combustion air, and something still needs to reduce those modeled losses by about 80% (whether that's through excellent insulation and air sealing and an ERV or any combination of the above - the net result still needs to be an 80% reduction).

Given how common oversizing of hydronic systems is, and the average age of housing with hydronic heating, it seems like better controls have a lot of potential to improve things.