Actual gas usage or heat-loss calculation?

gasfolk

Can't do it twice. Sorry for the long post, but hope you can steer me right.

3200 Sq. ft., brick house in New England, with 1100 sq. ft. of basement (heated by jacket losses from WM EGH-85), and with tons of CI radiators. Return water temps of 90 to 100 most days—rarely up to 118 max with a delta T of about 5 (converted gravity, two-pipe system, now with Tekmar 352 PS piping).

I ran heat-loss calculations with two ACCA Manual-J-compliant programs. Results: 80 MBtus/hr.

Plugged in monthly gas usage and HDDs from bills over past seven years. Picked the month with the highest Btus/HDD and calculated the Btu/Hr heat load for a design day of 5 degrees (99.6% heating DB) outdoors and 65 degrees indoors. Assumed the EGH-85 (350 MBtu input, 288 MBtu output, steady-state thermal efficiency of 75.6%) would have an efficiency around 50% when running at a 25% load. (See raw data below).

From all this, I get a calculated design-day heat loss of 44,543 Btu/hr, which makes me think Munchkin T50M (or similar?) and maybe indirect DHW. Let me add, no allowance was made for the gas stovetop and water heater, and this “high Btu/HDD” month occurred before a change from R-19 fiberglass to R-27 Icynene in the cathedralized roof, with extension of the new insulation over several large, previously-un-insulated roof areas. With the new insulation, gas use appears to be down at least 20%. ALSO, if necessary 1)insulation could be added to the basement and a crawlspace, 2)closure of the chimney flues (3 of them) could be improved, and in a pinch, 3)existing electric baseboards in the attic and basement could be cranked.

QUESTION: Have I miscalculated or taken a wrong turn? Should I go with the software (80 Mbtus/hr) or actual gas usage (44 Mbtus/hr)? Should I add in an additional safety factor? Are there other issues to consider? Thanks for any advice you may offer. gf

Raw data (for Feb 2003, the worst Btu/HDD in seven years):

Total gas ccf: 424
Summer average cooking & DHW gas: 19 ccf (Probably significantly higher in winter because of cookouts in summer and longer, hotter showers in winter?)
HDD: 1222
Therm factor (per gas bill) 1.027 (i.e. Btu/ccf: 102,700)
Design temps: outdoor 10 F, indoor 65 F
EGH-85 Thermal efficiency: 75.6%, guesstimated EGH-85 efficiency at 25% load: 50%?

EBEBRATT-Ed

Well you have a lot of data. I would use a heat loss formula you feel comfortable with and go with that.Or tra a few and average them. There is a SLANT FINN program available free on this web site.

Going off gas consumption you have to assume the efficiency of the existing boiler which may or may not be correct.

Personally I would go with the heat loss programs the more specific information it asks for the more accurate it will be. Some "simplified" estimates will tend to be to conservative (equipment will be to large)

Just my opinion=--I'm sure others will follow

ED

jerry scharf_2

I'd design for 80MBH

You are making a bunch of assumptions on the efficiency of the boiler in doing the gas based assumption, so I consider that less accurate. But heat loss calcs aren't perfect. They tend to take some wild guess at infiltration (or force you to.) Thay also assume your insulation is fiberglss, with all the real world derating it needs. For the Icynene, these are incorrect.

But when I get to the practicality of it, I would design for the 80MBH. It is very hard to find boilers that modulate lower than about 15MBH, so I see little advantage in going with the smaller unit if you are close. Since you mention the Munchkin wall hungs, they both have the same bottom firing limit. The T50 has smaller steps between low fire and high fire and is shallower. Certainly not enough to make it worth the risk in my book. If you are using the vision 1 and an indirect, the T80 can heat more water as well.

jerry

gasfolk

Boiler efficiency calculation? Multiple Boilers?

Thanks for your good thoughts. I am grateful for the chance to bounce this off of thoughtful folks.

The T80M would address that issue. Is there any way to calculate the efficiency of a boiler running at partial load? I have seen an equation and will try to find the reference and post it here.

Just had a suggestion to ADD a WM CGa-25 (Input 52 MBtu, Output 44 MBtu), keep the EGH-85 for backup and for coldest bin hours, and use MBS (Floyd, any thoughts?). Plenty of room in the boiler room.

gf

jeff_51

one more thing to consider

and that is liability. If you use the industry standard for heat loss, wheather I.B.R. or manual J you've covered you rear as long as you no's are correct. WE all want to be as efficient as we can for our custs and our own pride at doing a top notch job. We all want to experimant to see what we can really do. If you can get a cust who's willing to work with you and go with the lower numbers with the possiability of adding the 2nd unit later, your cust might get a much more efficient system. Personaly I think alot of our stuff is oversized. Try sitting down with the cust, lay out both of the options and see if he will work with you. Good job to practice you're sales techniques. That gas usage, if all info is correct tells a very differant story than the heat loss calcs doesn't it? Keep us posted

jerry scharf_2

my $.02

The efficiency of the boiler does not really change wih modulation. In a condensing boiler, it changes with return temps. So during the shoulder season, I expect higher efficiency than during the design days. It's the opposite of a cast iron boiler!

As for 2 boilers to get 50-80 MBH, that seems way too complex. When you are done with the controls and piping, you have invested a great deal of effort and you are still running with a lower efficiency boiler. Also the fixed fire boilers will be starting and stopping on all but design days, which increases the difference in efficiency from the T80...

jerry

Mike T., Swampeast MO

My gas consumption vs degree days with a condensing/modulating boiler confirms that. When comparing new boiler to old (plain cast iron very oversized) the old [seems] to have become significantly more efficient in colder weather despite the oversizing. Shoulder season was the absolute killer for the old but it's where the new excels.

gasfolk

Partial-Load Boiler Efficiency Equation

In trying to adjust the current gas usage (which reflects input Btus) to obtain an estimate of the house's heat loss (output Btus), I needed to estimate the apparent efficiency of the EGH-85 in its current installation. Here is the equation I used for conventional boiler efficiency at different loads:

E = Emax (1 – e ^ (-4.9L))

Where
Emax = Optimum efficiency,
L = Load factor = actual output divided by maximum output,
e = Base of natural logarithm = 2.718…

(Source: Manczyk, Henry, C.P.E., C.E.M., in “Optimal Boiler Size and its Relationship to Seasonal Efficiency,” Dec 2001, www.energy.rochester.edu/efficiency/optimal_boiler_size.pdf, page 29.)

This equation suggests that at loads of 0.6 and above the apparent efficiency is 95% (or higher) of the Emax. On the low side, as the load drops below 0.3 the apparent efficiency plunges. Nice that This fits with our shoulder season experiences.

If I assume a maximum heat load of 80 Mbtu/hr, the Load factor (L) for the EGH-85 would be

L = 80 Mbtu/hr divided by 288 Mbtu/hr = 0.278

The EGH-85 has a steady-state thermal efficiency of 75.6% = Emax

Plugging in the variables gives a partial-load EGH-85 efficiency of

E = 75.6% (1 – e ^ (-4.9*0.278)) = 75.6% (0.74) = 56%

So, if I accept the 56% apparent efficiency, I still get a design heat load of 47.7 Mbtu/hr using PRE-Icynene gas consumption. With POST-Icynene data (February 2005, 353ccf, 1150HDD) and an efficiency of 56% gives a design-day (10 degrees outside) heat load of 40.5 Mbtu/hr, and that is with NO adjustment for gas dryer, DHW, and stovetop.

HERE IS THE PROBLEM: If the POST-Icynene heat load is indeed closer to 40 Mbtu/hr than 80 Mbtu/hr, then the EGH-85 load factor (L) is closer to 40 / 288 = 0.14, and the above equation yields a partial boiler efficiency of 37%. At that apparent efficiency the calculated POST-Icynene heat load of the house drops to 26.7 Mbtu/hr.

I totally agree that the boiler efficiency requires an assumption, but am I wrong in thinking that the number of assumptions in calculating from the gas usage are fewer and more tangible than the unknown assumptions in the excellent heat-loss software packages I am using. It would be nice to know how much faith to put in this part-load-efficiency equation regarding a boiler that may now (post-icynene)be 10 times (yikes!) the required size. How crazy is a predicted efficiency of 56% at a 27.8% (0.278) load factor? Or 37% at a 0.14 load factor?

I have to thank you all again and again for your insights and for helping me think this through. gf

Chuckles_3

I haven't checked your calculations, but I can address the general principle. Those much-touted heat loss calculations are only as good as the assumptions that go into them. Garbage in, garbage out. Our house has actual usage of about 80MBTU on a design day. Four contractors did heat loss calcs and came up with numbers in the range 150-200MBTU.

If you have done your actual usage calculations correctly and added some cushion, you are doing better than any heat loss calc.

The business about the T50 vs T80 is also correct...there is no reason not to go with the 80. The price difference is pretty small.

S. Orbine

my two cents;

Others have pointed out that present actual boiler efficiency is an unknown, but I'd bet that it isn't alot higher than advertised (lower maybe, not higher) which is what it would take to make the actual usage low.

I'd say the empirical data is far superior than any calculated estimate. Further, I believe there to be a fairly large margin (overestimate) built into the manual J calculations.

Unless your gas meter is way off (or your thermometer), use the actual data and add whatever cushion makes you feel comfortable.

gasfolk

Actual gas usage

All of your comments are eye-opening. Before I thought to ask for records of gas usage, the 80 Mbtu/hr heat-loss calculation had me leaning toward something like the Munchkin 140M—“just to be safe.” Of course, the 140M only modulates down to 43 Mbtu/hr, so if the calculations from actual gas usage are reliable, that 43 Mbtu/hr is higher than the design-day heat load. In that house the 140M would have to cycle on and off for most of the heating season.

Today you are right, the T80 makes clear sense over the T50. But if you guys can offer (as you are doing) support for the 40 (or less) MBtu/hr heat load on this big old house, then HTP and WM and Burnham and the others should start making smaller boilers with higher turndown ratios. Maybe we are being saved more than we know by these modulating boilers. Maybe the dearth of boilers at the small end is because we keep ordering based on inflated Manual-J estimates. Maybe I’ll recommend holding out till someone makes a 40 MBtu/hr boiler with a ten-to-one turndown ratio. But when will that happen? Does anyone know whether five-to-one coming to the T50?

Actual gas usage. Very interesting. Can anyone support or criticize that partial-load efficiency equation?

Thanks, gf

jerry scharf_2

talked to the folks at HTP about this

gf,

Brian French at HTP is one smart guy (there's a smart Guy there as well.) Both the T50 and T80 have the same Duns gas valve. The problem is getting stable combustion below 15MBH. Even if he can do it in the lab, what is the chance it will work reliably in the field? His exact comment to me was "Sure I could do that, but you wouldn't like the cover rattling off."

Now Brian showed me some things that might make that work, and I may check again later on this. But the other thing is since there is not much efficiency penalty for start and stop on a condensing boiler, cycling it a few times an hour in the shoulder season shouldn't be too bad. It seemed like there was not too much demand perceived to push that end of the envelope.

jerry