End of Season Report V4: Thermal Purge-O-Rama!

fentonc

With this year's heating season done, I finally got a chance to sit down and analyze this winter's boiler performance (for previous seasons, see Part 1, Part 2 and Part 3). There were a few differences from previous seasons:

• Zone synchronization is gone. Too weird for too little gain.

• After reading about a similar scheme, I attempted a dedicated 'zone protection' scheme, where zones circulated for at least 5 minutes every 45 minutes, whether they needed heat or not.

• The upgraded Boilertron was re-wired to implement a thermal purge scheme following heating cycles. After extensive simulation in the fall, I settled on a scheme where, following all calls for heat being satisfied, the burner was inhibited and all three zone valves were opened up until the water temperature dropped to room temperature.

• Just to make reasoning about the behavior easier, I disabled thermostat setbacks. The 2nd floor was maintained at 65F, the main floor was set at 68F and the basement remained at 62F (like last season). Previously, both upper floors had setbacks to 62F over night and during the day when the house was unoccupied (heating to 65F and 68F during morning/evening hours). The effect was to raise the average temperature of the house significantly compared to previous seasons, so I did expect some amount of increased fuel usage from this, hopefully countered somewhat by the thermal purge scheme.

How did I do? Mixed results, but overall pretty well. No regrets on getting rid of zone synchronization - it really was too weird for everyone in the house. The zone protection scheme was also a disaster and lasted about a week - it led to significant overheating while somehow simultaneously wrecking the efficiency. Good riddance!

I was very satisfied overall with the thermal purge scheme - any overheating was very minor and went unnoticed (unlike with the 'protection' scheme) and the improvement in efficiency (estimated delivered BTUs vs consumed fuel) was significant, especially at very low duty cycle. Overall, as predicted, fuel consumption per HDD65 appeared to be slightly higher, but the house averaged several degrees warmer.

And now, for the charts:

As in previous years, burns were dominated by the ~90 second recurring burns when only 1 zone was enabled, although in aggregate about 1/3 of the burns were 4 minutes or longer.

An interesting side effect of the purging scheme is that it very quickly became a 'continuous' system. In the shoulder months, frequently the single initial burn would satisfy the thermostat, and it would take 7-ish minutes to purge the remaining heat, but as it got colder a new call for heat would occur before the previous one had finished purging, so the system ran nearly continuously for the whole winter (the chart below just merges everything longer than 30min).

In terms of fuel usage vs outdoor temperature, things were slightly higher due to the higher average indoor temperatures:

And, at least as-measured, the efficiency was significantly improved vs previous seasons, with the boiler hitting it's thermal efficiency of about 75% (tested in October) across pretty much the entire duty cycle range. The efficiency is estimated by tracking the average zone temperature for all 3 zones, and integrating every 1 second all season (same as previous seasons). The boiler should have a nominal thermal efficiency of 84%, but the boiler only fires at 120KBTU/hr vs its nominal 140KBTU/hr due to low gas pressure. This is something I'm hoping to fix this summer. Anyway, shout out to @Roger , as Energy Kinetics was my inspiration for the thermal purging scheme. If you've got an oversized boiler, it works great!

Steamhead

Nice work! I would have kept the setback though, since that would eliminate one variable.

Now we need to come up with a multi-zone control that incorporates the post-purge feature. Caleffi, Taco, Argo et al, ya listening?

For hot-water systems with just one zone, the Beckett AquaSmart control has a thermal post-purge feature you can enable. I don't see a way to wire it for a multi-zone setup though.

fentonc

I'm hoping to get the gas pressure issue fixed this summer and then add the setbacks back in next season, now that I'm pretty confident the all-zone purge scheme will handle the 'freeze protection' issue pretty well (even with setbacks on 2/3 of the zones). I haven't thought about it carefully, but it's probably possible to do the all-zone thermal post-purge with just some relays. It really works better than one might expect (at least in my house), due to the following:

• There's only about ~10k BTUs stored in the boiler after firing (for my 140KBTU/hr CI boiler)
• Any zones that haven't been active in a while will be filled with cool water, which can mix with and absorb the BTUs very quickly
• The overheating is minimal when the BTUs are distributed evenly throughout the house, and the inactive zones will typically have been a few BTUs below their setpoint anyway
• For the inactive zones, it will mostly just delay them calling for heat again for a bit longer
• It can very quickly cool the boiler down, virtually eliminating standby losses for oversized boilers

RascalOrnery

I did a simple post purge with my boiler, since it is coal it can overshoot pretty hard after a heat call . This makes wasted idle heat. I just put a single acting aquastat (NO) in the primary loop, just below the high limit setting, in series with the zone circulator.

Once the heat call is ended, a delay off timer receives power. It is set to fifteen minutes, it sends power to the circulator still through the aquastat and finishes five degrees above my low limit setting (I have a tankless coil setup)

works well to keep idle temps as low as possible (though not as low as yours)

https://coalpail.com/coal-forum/viewtopic.php?p=860702#p860702

hot_rod

Wow lots a data to sort through!

On a design day, how long is the boiler run cycle? If the boiler fires before the 7 minute post purge ends, seems the short cycles are still present?

Also if you pull the boiler to room temperature every time, does it warm beyond condensing mode when it fires, for a single zone call for example?

Those are the two concerns with over-sized boilers, short run/ off cycles and the boiler runs too cold, as it covers the load before the return warms adequately.

As I recall one goal was the de-thunking of the zone valves frequently, did that get solved?

fentonc

On a design day, the boiler burns for about a 25% duty cycle for the whole day (I think I've encountered about 7 design days in the last 4 years, typically 1-2 a year). The longest recorded burn of this past season was 1,007 seconds (so nearly 17 minutes) - that must have been a cold start where all 3 zones called for heat at nearly the same time. Typically, one zone will call for heat from cold start (so boiler is at ~65F), the boiler will burn until it hits the 180F aquastat high limit, it will drop to 160F and burn for ~85s before it hits the high-limit again, repeat a couple of times and then start purging.

The boiler pretty much hits the 180F high-limit every time it fires, whether cold-starting or maintaining temp during a call for heat. It's just a low-mass system with a boiler that's almost always way oversized (and occasionally, only somewhat oversized). I've never seen a hint of condensation anywhere in the system. It definitely short-cycles the burner like crazy though - probably 90% of the times the boiler has fired in its 9 year lifetime have been for ~90 seconds.

Good memory about the de-thunking! I did sacrifice that feature to use the 4th relay to inhibit the burner rather than the circulator. I think the near-continuous running helped mitigate that at least somewhat, as I know i heard it a handful of times this season, but it wasn't enough to be annoying.

fentonc

One fun thing I haven't analyzed yet - I recorded the per-zone delivered-BTUs and temperature for every hour of the season, so I should have extremely fine-grained heat loss data for my house that I can update my simulator with.

Steamhead

@fentonc , what model aquastat relay does your boiler have? Does it have outdoor reset? ODR or an upgraded aquastat relay (Aquasmart or HydroStat) would eliminate much of the boiler overheating when only one zone calls.

fentonc

@Steamhead - it's a "Model 1135-605" that came with my WM boiler. In theory it does support outdoor reset, but my simulation modeling seemed to indicate that the aggressive thermal purge I implemented significantly outperformed just running at a lower water temperature, as it achieves a significantly lower average temperature at the boiler. Running at a lower water temp does significantly outperform running at a much higher water temp when there is no thermal purging due to the reduced standby losses, but otherwise the longer runtimes mean the boiler stays at an elevated temperature for longer.

My setup is a weird 'hybrid' aquastat, as I left the WM aquastat in place, but I have a custom PLC that sits between the 3 thermostats and the zone valves, and can also inhibit the burner on the boiler while keeping the circulator running. I'm definitely open to trying new control schemes, but I think the one I settled on works quite well in terms of achieving high efficiency.

Steamhead

Here's the outdoor sensor- it's not expensive:

https://www.supplyhouse.com/Weil-Mclain-381-356-586-Outdoor-Temperature-Sensor-Kit?_br_psugg_q=381-356-586

Can't find the manual online quickly, but I'd certainly be interested to know how it performs when added to your current setup.

fentonc

The sensor isn't expensive, but there don't seem to be an programming instructions in the manual (maybe the only configuration is based on setting the aquastat high limit?).

I have a strap-on supply water temperature sensor to implement the thermal purging, which allows my PLC to effectively override the aquastat high limit (as long as it's below the limit enforced by the WM aquastat) whenever I want. Given that my system can actually provide enough heat with ~120F SWT, I was somewhat worried about condensation if I left the SWT too low. A scheme I had considered was to try to do an "Indoor Reset" - because I can measure how quickly the water temp is dropping (and I know how much radiation is connected), I can infer the near-real-time actual heat loss, and calculate what average SWT would be needed to meet that. To avoid condensation, I could leave the high-limit at 180F or so, but then modulate the average SWT by controlling the low-limit before the burner kicks back on. Maybe on the first burn, I let it get up to 180F and then fall back down to 100F. On the 2nd burn, I only let it fall down to 105F. 3rd burn, 110F, etc.

I don't currently have a way for my PLC to read the actual outdoor temperature like an ODR sensor, but I think I can infer the same info if it gave me something in terms of efficiency. All that being said, that last chart I posted seems to show the thermal purging basically lets me hit the nominal boiler efficiency at practically 0% load.

fentonc

This season I recorded about 2000 hours of hourly 'delivered BTUs' data, along with the outdoor temperature, which gives me this nice chart:

Delivered-BTUs are calculated using a 3-hour moving average. I have quite a few SW-facing windows, so it's got to get below 20F outside before I completely stop seeing long periods with the heating system shut off. The actual outdoor temperature is probably 2-3 degrees below what I measured, as the sensor is attached to the back wall of my house.

Roger

@fentonc , I'm glad Energy Kinetics could be your inspiration for thermal purge. It's one element that can help minimize idle loss (along with low mass, heavy insulation, and other design features). All the best strategies combined result in boiler operation near peak efficiency from small part loads up through steady state operation, in other words, oversizing effects can be nearly eliminated. I think you'll find success with night setback as well.

Best,

Roger