Boiler modulation question

[MikeG88]

The 2.6 GPM at 77F rise was the one major negative I picked up on from the brochure. It is a 2 bathroom house but I currently occupy it alone, and am not in the habit of running the other water using appliances while showering. Another adult would be fine, but children would find it hard to resist using a second shower at the same time. I guess that's the one place the Navien really shines- I can open multiple faucets and it keeps up fine.

I see it is about 5% more efficient than the Navien but that alone probably doesn't add up to much savings over the Navien. I wonder if with it's much lower fire rate I could run it at lower temperatures -closer to/within the condensing zone, therefore generating some savings there? If I did replace the Navien my main reason is to get something I feel would be more reliable and less noisy, but being able to recoup some of the cost (over time) through savings would help justify the expense.
Is there another boiler I should be comparing to the Lochinvar NKC110, especially asfar as the minimum firing rate?

 

[Dana]

Unfortunately this is THE problem with l0w-mass combi-boilers. The minimum fire is (far more often than not) too high to operating in condensing mode without excessive cycling, if sized reasonably for the domestic hot water load. They are a better match for houses with modest to moderate hot water needs and large space heating loads, which is pretty far from the description of the average US house, and almost the exact opposite description of your house. Ideally with 54' of baseboard you'd want a boiler with a min-fire output <10,000 BTU/hr to get decent condensing efficiency out of it, and with 2 bathrooms a high fire of 180,000 BTU/hr or higher.

A combi boiler with a 20:1 turn-down ratio would work, but according to the tooth fairy those boilers are too heavy for his unicorn to carry.

When the radiation output at condensing temps is well below the minimum firing rate, the alternatives are a tank-type combi-boiler ( or cobbling-up a self-designed combi around a condensing tank water heater), or a buffer tank, or a high-mass hydraulic separator to limit the cycling when operated at condensing temperatures. Any of those run into real money.

Rather than pouring the money into making an ill-fitting combi boiler work better, in a better than code-R house like yours and CT style electricity and propane pricing there's probably better financial & comfort return by adding a 1- ton ductless mini-split to carry most of the heating load (all of it except during cold snaps) which will also cool / dehumidify the place at high efficiency during the summer. A Fujitsu -12RLS3H or Mitsubishi -FH12NA puts out 14-16,000 BTU/hr @ +17F (one of the temperatures at which efficiency is tested). When it's under 10F outside the difference in operating cost between a mini-split and propane boiler isn't very big, but at temps of 25F and higher the mini-split's as-used efficiency soars. There are 1.25 ton versions of each that would pretty cover 100% of the space heating load, but when it's +5F outside the room to room temperature differences might become too large. If you're ever considering that option seriously we can fine tune it.

 

[Sponsor]

 

[MikeG88]

Thanks Dana,

I actually do have a 2 1/2 ton Mitsubishi Mini-split system. There are 2 9k BTU units in the living area ( one in the master bedroom and one in the living room that opens to the dining and kitchen area). and a single 12k BTU unit in the basement. I don't usually run the unit in the basement, but I have used it to to cool/dry the basement and if I did have to use the system to heat during single digit and below temps- it's nice to know the mechanicals etc in the basement would be safe and the heat would rise to the living area. I've used the two 9k units to heat in the fall season but do have a couple small unoccupied bedrooms that I use for home office and storage, and a small laundry room, all with base board heat but with the doors normally closed and heating with the Mitsubishi, these rooms do drop noticeably below the rest of the house. I have used the system down to 20F, but these rooms did drop about 10F below the rest of the house. I would probably have to add some supplemental heat to the bed rooms if they were going to be occupied with the doors closed, or find louvered doors that match the rest.

I guess I should use the mini-split system during the mild (short cycle) periods and use the boiler during the rest of the winter. My original intention was to get air conditioning but the ability to heat (especially as a capable back up even below 0 with the hyper heat) was an additional point. I guess I am looking for a one size fits all solution but they don't yet make the boiler I need.

 

[Dana]

Use the minisplits ALL the time, not just the shoulder seasons. Even if it doesn't have sufficient capacity to keep up with the load on it's own, the efficiency of the MXZ-3C30 is at least twice that of the electric boiler even at +5F. At -10F it's still at least 1.5x as efficient, using only 2/3 the amount of power per BTU as the boiler. The hydronic system can be your primary backup for cold snaps or when you need to heat the place up quickly, but the mini-splits are gonig to cost only about 1/3 as much to run on a seasonal average, 1/4 as much when it's in the 40s or higher. The tested HSPF efficiency of 10.0 means the seasonal average will put out 2.93 x as much heat per kwh as the boiler, but without the distribution losses to the basement it'll be a bit more than 3x as much, as a system.

A safe small oil-filled radiator type space heater for the temporary-use doored off office/bedroom spaces is also somewhat more efficient than the electric boiler, since it doesn't have the distribution losses to the uninsulated basement.

Air sealing and insulating the basement is probably the next step then.

[edit to add]

I was getting this thread confused with another thread of a house in CT using an electric boiler. (Need to drink more coffee prior to posting, I guess. )

Run the cost numbers per million BTU assuming 7000 BTU/kwh for when it's single-digits outside against 80,000 BTU/gallon propane (87% efficiency, non condensing) to see where the financial crossover is.

[further edited to add]

CT residential propane is running about $3/gallon right now, electricity is 21 cents/kwh.

1,000,000/80,000= 12.5 gallons/MMBTU, which at $3/gallon costs $37.50/MMBTU

1,000,000/7,000= 143 kwh/MMBTU which at $0.21/kwh costs $30.00/MMBTU

So even when it's +10F outside heat from the ductless system is ~20% cheaper than propane at near-condensing water temps, and would still be 10% cheaper than 95% efficiency condensing propane.

To get the best efficiency out of the ductless system a "set and forget" approach works better than overnight setbacks. Energy savings due to the lower heat loss at a lower indoor temperature gets more than eaten up by the lower efficiency of having to run full-bore during the recovery ramp. Bumping it up or down a degree or two at a time as-needed doesn't hurt much, but 5F steps (or greater) take a toll.

 

[MikeG88]

Thanks for the math showing where the efficiency comes from. Most of my experience heating with the minisplits is when the outdoor temps are in the 30's or higher. The seem to run a lot of the time but not pushing a whole lot of volume. The seem to start off a little above room temperature and slowly increase until the cycle is complete. As per the manual recommendation (they are MSZ-GL09NA units) when running multiple units in heat mode- I have the fan speeds set to auto. Of course I can increase the fan speed and bump the set point way up and get a lot more heat and air flow, but I assume the lower temp, lower volume and long runs is actually the unit modulating to be efficient rather than "impress" me with a wave of tropical wind. I've observed that they seem to operate in tandem, one joining the other during a cycle (or sometimes starting together) and both ending at the same time.

Another thing I've observed that when outdoor temps get to 30 or so (and below) that both green lights (rather than just the one always on) will sometimes come on. The manual says that means the room temperature is "about 4 or more from set temperature" but per thermometers in the rooms the temperature is still fairly close to set point. The single green light means the temp is "about 2 to 4F from set temperature" but that light never goes off as long as the unit is powered. I'm thinking the lights have more to do with how hard the unit has to work to maintain temperature rather than how far from actual set point it is, unless the thermostat in the unit is reading something different.

 

[Dana]

The temperatures are sensed at the incoming air at the heads, and it's common to have significant offsets between the sensed temperature and the average room temperature, especially when the heating loads are higher. That's fine as long as you're not hung up on what the number says, and just bump it up or down a degree or two at a time to stay comfortable. If it bugs you, Mitsubishi does sell both wired and RF-link wall thermostats. (Most people don't bother.)

The minimum modulated output of the compressor is much higher than the minimum output of any single head, and when one zone has passed it's setpoint it forces the other heads to run a bit faster to take more refrigerant. The heads control how much refrigerant is being called from the compressor, but they must be running some algorithm by which they spin up another head to run more efficiently rather than forcing one head to take the full minimum output of the compressor (which for Mitsubishi units is roughly the maximum output of a single half-ton head) , which is less efficient than running two heads a minimum speed. Worse still would be cycling the compressor on/off frequently. Ideally they would run at or near minimum speed most of the heating season, cycling very occasionally, only ramping up when there's more load. The minimum modulated output of a single GL09 is 4500 BTU/hr @ 47F, but the compressor's output is more like 7000 BTU/hr, so there's some decision tree they're chasing as to whether to overheat the other zone a little bit to have sufficient load for the compressor vs. turn off early and under-heat a zone.

 

[Ron Beck]

A few questions here about this conversation. Why would we be satisfied with a two to three minute run time on a mod/con boiler? I would rather see 15 minutes or longer. If the cycle is off every three minutes that is telling me the boiler is still inputting much more fuel than required. If we reduce the input would it not run longer?
When the boiler cycles off our stand by losses increase dramatically due to pre-purge and post purge cycles. More cycles is harder on the equipment. More ignition cycles is harder on electrodes. More cycles on the gas valve solenoids.
Isn't modulation more important in operation cost than condensating? The lower the fan runs the ratio of fuel input to fire side heating surface increases as fan slows down driving efficiency up.
When running low enough water temps to condense we are only removing enough heat from the flue gasses to cause the condensation. We are only recovering about 9% of the btu's from the flue gasses which is not that many.

 

[Dana]

Of course 15 minute burns is easier on the equipment than 3 minutes, but at 3 minutes the ratio of BTUs blown away during a flue purge & ignition cycle to the total heat of BTUs generated by the burn is down to a low single digit percentage, and you'll (usually) be well under 10 burns per hour.

If you lose 1-3% in cycling, but gain 5-9% in condensing, it's usually "worth it". If it's cycling a lot, losing 5% to cycling and only gaining 5% in condensing it's just abusing the equipment. Without a lot of instrumentation and measurement it's hard to nail down the exact crossover point, but at 5 burns per hour it's very hard on the system at all (even though 2 burns per hour is easier still.)

 

[Ron Beck]

I guess the difference is if you are above the approximate condensing temp or not. I always try to get the longest run time possible. During the rally cold snap a month ago it was surprising how many calls I received where the homes were not maintaining thermostat set point. I had the reduce the fan speed to lower input and the boiler. Stayed on and heated the homes.
My theory is on a demand if possible boiler should not shut off on temp but only when thermostat satisfies. When contractors call and tell me they are running 30 minutes I always say the same thing. Why not 45 minutes or more. If they say 60 minutes I say why not 60 minutes.
Maybe I could convince engineering to do some real testing on this. The 10:1 boilers does make this a bit easier than the 5:1 turndown.
I have learned most mod/con boilers do not operate in the condensing mode. Many do not ven gave the OD sensor installed and are 100% or more oversized. We need more heat loss, less sizing off the old boiler or sizing off the connected load.

 

[Jadnashua]

Properly sizing a boiler actually takes time and skill. Not everyone is both capable and willing to expend it. A discerning user and a competent designer does make a difference. Many people just look at the cost (a good analysis takes time, thus money). It's far easier to just say, well, if a 200K BTU boiler kept your house warm, we'll just put in another one and leave it at that. Trying to sell long-term comfort and economy is a harder sell than just replacing with the same.

The ideal situation would be to adjust the heating unit's output to exactly the heat loss and have it run constantly. That won't happen.

 

[Ron Beck]

You are absolutely correct. It will never run long cycles if the time is not taken to do our jobs correctly. The boiler will never turn down far enough. The closer we size the boiler to the correct size the lower it goes than being 1, 2 or even three times oversized. Granted it does help to reduce fan input to a lower rpm.changing the fan rpm' reduces the turndown but to me that is better than running shorter cycles.

 

[NY_Rob]

You are absolutely correct. It will never run long cycles if the time is not taken to do our jobs correctly. The boiler will never turn down far enough. The closer we size the boiler to the correct size the lower it goes than being 1, 2 or even three times oversized. Granted it does help to reduce fan input to a lower rpm.changing the fan rpm' reduces the turndown but to me that is better than running shorter cycles.

When retrofitting a mod-con into a system that previously had a 180F SWT boiler... unless you add radiation you will most likely not be using it to peak efficiency.
Even with radiation overdesign as was done in the past- you just don't get enough BTU's out of a ten ft section of fintube running it at 120F SWT to heat a room that was designed to have 180F water pumping through that original ten ft section of fintube. So, the room is cold, the installer gets called back and cranks the ODR curve (or disables ODR completely) to give 180F water at 40F outdoor temp and everyone's happy.

10ft fintube at 170F AWT = 5,100 BTU's
10ft fintube at 120F AWT = 2,100 BTU's

I had to do quite a few modifications to my existing radiation be able to use 114F AWT at 32F outdoors temp.

 

[Dana]

I guess the difference is if you are above the approximate condensing temp or not. I always try to get the longest run time possible. During the rally cold snap a month ago it was surprising how many calls I received where the homes were not maintaining thermostat set point. I had the reduce the fan speed to lower input and the boiler. Stayed on and heated the homes.
My theory is on a demand if possible boiler should not shut off on temp but only when thermostat satisfies. When contractors call and tell me they are running 30 minutes I always say the same thing. Why not 45 minutes or more. If they say 60 minutes I say why not 60 minutes.
Maybe I could convince engineering to do some real testing on this. The 10:1 boilers does make this a bit easier than the 5:1 turndown.
I have learned most mod/con boilers do not operate in the condensing mode. Many do not ven gave the OD sensor installed and are 100% or more oversized. We need more heat loss, less sizing off the old boiler or sizing off the connected load.

Exactly.

Morgan Audetat (aka BadgerBoilerMN) has one house set up where the ODR is so finely tuned he's running, without a thermostat, a hard-wired continuous call for heat. Why only 60 minute burns- why not 60 days?

Why only 60 day burns instead of 120 days?

At some point it hardly matters, but if you're on the edge of squeaking some condensing efficiency out of it at the expense of modest cycling, you have to draw a line somewhere, even if it's a bit of a WAG.

My personal non-scientific not well modeled rule of thumb is >3 minute burn times and <10 burns per hour, if that's what it takes to buy some condensing efficiency. But it's really just a WAG- I'm sure the exact crossover point on efficiency and lifecycle savings will vary from system to system, boiler to boiler.

MikeG88's radiation won't emit the full minimum fire output of his boiler at condensing temps, and would need an AWT of 145F or so to balance well enough for very long burns. To get condensing efficiency he needs to lower the temperature and let it cycle a bit, the question is/was how much cycling is tolerable, and at what point it becomes boiler-abuse.

 

[MikeG88]

Dana I took your advice and have been heating with the minisplits for several days (the two GL09s). I did observe last night (with outside temps in the mid 30s after a day that went up into the 40s) that I was getting heat cycles based on how long the vanes were in heat position) for 10 or 11 minutes with maybe 4 or 5 minutes with the vanes up in-between. (I believe I read the enforced compressor off time is a minimum of 3 minutes, so counting the time the compressor runs before the vanes open, this was probably the minimum.) The air flow and temperature seemed pretty typical. This morning (around 32F outside the heat cycles seemed longer as well as the off time but I didn't get a chance to time them). Tonight at 39F outside (another day that was in the 40s) I timed a heat times of 25 minutes with 7 minutes in-between but after a few cycles the times seem to be shortening now down to 14 minutes of heat with 6 minutes off (still the same 39F outside). Is this normal? The head in the living room (open to kitchen/dining area and hall to closed-off bedrooms) is usually running during the complete cycle and the unit in master bedroom joins in some of the cycle. Fan-speed, on both, is set to auto as the manual recommends for running multiple heads together. I was wondering if there was some point where it was more efficient to run just one head but re-reading your earlier response it seems it is best to run both heads together to get the best out of modulation (especially since the compressor puts out more than the minimum of one head)? This doesn't seem to be a case of the system having trouble keeping up, (even when both green lights come on the heads don't seem to be really working that much harder to catch up), neither head is acting like it would if I turned the heat up to the 80s, more like it's running closer to minimum output, it seems to be maintaining (or getting back to) set point without the excessive overshoot it sometimes gets when it's warmer out, but why wouldn't it now overshoot set point a bit on purpose so it could take a longer than minimum break between heat runs?

Would running just one head when cooling (especially during more mild times) get even more de-humidification as unit would work harder and/or longer, but would I be losing efficiency? In cooling I keep the bedroom unit (as it is the smaller zone) set a degree or so lower than the living to get it to work a little more. I also have it set a degree higher for heat mode.

I did look up the external thermostats but none were inexpensive (and would need one for each head). I did hope to find a humidistat but didn't notice one. That would be nice for the basement unit as I had hoped to use it to dry the basement, but found that in dry mode it seems to run non-stop, even when the other units are not being called on to cool. I assumed it was cheaper to run my stand alone dehumidifier that shuts off when it achieves the desired humidity, rather than run the compressor constantly. It was recommended, to me to set the basement unit to cool the basement to a few degrees below the temperature it would be without the cooling, to get it to reduce humidity in the basement?

Thanks!

 

[Dana]

The cycling seems pretty normal for the conditions. At recent days' outdoor dew point temperatures in southern New England defrost cycles are frequent since there's more water in the air than when it's colder, and the outdoor coil temps will be below freezing when it's within ~10F of freezing outdoors. Do the GL heads have an indicator light for when it's running in defrost?

The algorithms by which the head decides to speed up or back off are proprietary. I'm pretty sure there is at least some measurement of the rate of temperature change that keeps it from running flat-out at high speed immediately, but after a sustained period of too-slow temperature rise it would probably speed up, then start to back off as it approaches the set point rather than overshooting.

During the more sticky humid times it's best to still run both heads, but set them both to "DRY" mode. In DRY mode the SEER is lower, but the latent/sensible cooling ratio is higher (which is why the raw SEER efficiency is lower- it's a function of how SEER is tested.) It's still far more efficient than running a dehumidifier to lower the indoor humidity.

 

[MikeG88]

There is no light to tell when the GLs are running in defrost. I have checked the outdoor unit when running below freezing and never noticed any overt frost etc on the fins but maybe didn't look at the right time or place. I have suspected that sometimes when the two green lights come on (manual says it means it's about 4F or more from set point) it may actually mean it's running defrost, but when I did check the outside unit it seemed to be normal (not sure what would be different or what to look for besides frost?). The only times I see both green lights come on is when it's around freezing or lower (or when powering system on, or obviously when bumping up set point). Nobody has ever confirmed that's what is happening.

I thought the 10 minutes on 4 minutes off might be something wrong- but I guess I'm too sensitive to times after watching the Navien.

 

[Dana]

Yeah, when analyzing this stuff sometimes you start to wonder if you're turning into one of these guys.

Run cycles of 10 minutes are good enough to hit pretty close to it's steady state efficiency, especially if it's just the fan that's cycling between heating and defrost modes while the compressor keeps spinning almost uninterrupted. But even from a cold start spin-up on the compressor a 10 minute on cycle is fine.

 

[MikeG88]

LOL saw the movie and read the book.

I think it's a cold start, at least most of the time. Fan definitely stops,
I think compressor totally stops, and air is not warmer than normal when it starts blowing.

As long as not doing damage or really killing efficiency I guess I shouldn't worry too much...

 

[MikeG88]

How can I tell when it runs a defrost cycle?

 

[Dana]

I don't know- what does the manual say? (Which rev of the compressor is it, the MXZ-3C or is it -3B, or -3A?)

During defrost cycles there will be drips coming off the outdoor unit's coil, and (hopefully) out the bottom pan's drain. Inspecting the drain for ice-plugs occasionly on weeks when it doesn't get above freezing is prudent.