New furnace: is it too good?
We are redoing our lower level and hoped to do the right thing by getting a tankless water heater, new windows, insulation and energy efficient furnace & comfortnet thermostats.
Our home is on 2 levels, 1050 sq ft up and additional 500 sq ft living space below so we have it zoned. Our contractor told us to buy a 60K BTU furnace; we opted for 96% AFUE with two stage with variable blower; Goodman GMVC960603BN.
We get high winds from the vents; we thought that the variable blower & two stages meant that the unit would operate on a low flame and low fan (cfm) level but that doesn’t seem to be the case, the furnace starts out slow then ramps up to gale storm.
Is this thing installed correctly? Is it too big? Is there a setting we can adjust?
Our general contractors put this in, but there is only one that I can speak with because of language barriers.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
John,
We need a lot more information before we can help you. To start, we would need (1) to know your geographical location or climate zone, and (2) to know whether your contractor performed a Manual J load calculation.
Step one for any heating system design is to perform a load calculation. Here are links to three relevant articles:
Saving Energy With Manual J and Manual D
How to Perform a Heat-Loss Calculation — Part1
How to Perform a Heat-Loss Calculation — Part 2
Step two for the design of a forced-air heating system is to design the duct system. Here is a link to a relevant article:
All About Furnaces and Duct Systems
Step three would be to commission the system. Here is a link to a relevant article:
Residential Commissioning
There is a good chance that your contractor never performed steps one, two, or three, unfortunately. You should start by asking you contractor about these three steps. You may need to hire a home performance contractor to check your system if (as I suspect) your contractor doesn't understand the three steps I listed.
The specs indicate that there isn't that that much difference between high and low stages. But there is a CFM adjustment that would reduce the "high winds". Also make sure that the thermostat supports two stages. Adjust zone dampers so that they never close to less than ~2/3.
Hi Martin, and thanks for your reply.
I wouldn't be surprised if your suspicions about our contractor not completing those steps are right. My conversation with the furnace installer hired by our contractor was limited to "how big should it be? 60K BTU? 66K BTU?" He said "60". We were responsible for shopping for it. The contractor made a suggestion of a single stage, single speed blower system even though we stated we wanted the opposite so I'm not sure how familiar they are with improved technologies.
We are in San Francisco, which is climate zone 3 I believe, so 1500sq ft x 40BTU/sq ft; 60K to meet the "rule of thumb" referred to in the article.
I will see if they understand your 3 questions, but even so, it is what it is and I have the feeling I will have to hire a home performance contractor on this brand new system.
Thank you, Jon.
I didn't know that I should have looked at the difference between stage 1 and stage 2 so should've done more research there, I guess.
I have read about the different colored wiring that correlates to different blower speeds and will reach out to the furnace installer about that too.
That's interesting about the dampers not being closed more than 2/3. It makes sense to me and seems a route to consider given that we are expecting the furnace to sometimes heat the whole 1500 sq ft, most often the 1050 sq ft upstairs and sometimes only the 500 sq ft downstairs.
As you probably suspect, when the furnace is only heating the lower 500 sq ft zone it is loud and fast! Reducing that cfm by leaving the upstairs damper upstairs makes sense.
In a zoned system, you have a zone controller between the thermostats and the furnace. Even if you have two stage thermostats and a two stage furnace, the zone controller may only be capable of single stage logic. When you use a single stage controller with a two stage furnace, the furnace is usually programmed to run on low stage for about 12 minutes and then to switch to high stage until the heat call ends. The furnace has no idea how many zones it is feeding. If you have a two stage zone controller, it would have more sophisticated logic like using low stage when only one zone is calling for heat and high stage when both are.
Perhaps worth trying hard-wiring it for only first stage operation.
A typical 2x4 framed house with clear glass double-panes (or storms over single panes) with 1550' of all-above-grade conditioned space would typically come in around 22-28,000 BTU/hr @ 0F outdoors, 68F indoors. A 2x6/R20 house with U0.35 (no longer meets code performance north of US climate zone 4) would come in 19-24,000 BTU/hr.
If you have a fuel-use history you can fairly quickly check the actual heating load at your actual 99% outside design temperature, (which is about 39F, not 0F http://articles.extension.org/sites/default/files/7.%20Outdoor_Design_Conditions_508.pdf ) analyzing last winter's fuel use against heating degree-day data following these methods:
https://www.greenbuildingadvisor.com/blogs/dept/guest-blogs/out-old-new
Run the numbers even if that was the "before" picture of your window & insulation upgrades.
So is a 60,000 BTU/hr 2- stage the optimal furnace for your house?
I SERIOUSLY doubt it!
Odds are pretty good that even at it's lowest firing rate the output it's probably more than 3x oversized for your actual 99% heat load. A reasonably tight & insulated 2x4 house at 30F would have a heat load of MAYBE 15,000 BTU/hr, 20,000 BTU/hr if it's all single-panes, no storms, and even an uninsulated brick house would typically come in WELL under 40,000 BTU/hr. For a point of reference, I have ~2400' of antique sub-code 2x4 framed space above grade, ~1500' of insulated basement. That much condensing burner could heat my house down to a fairly bracing -25F to -30F, WAY below 0F!
If you have the old gas bills, DO run the fuel use against heating degree days, then report back. Or if the blog piece isn't clear enough I can walk you through it if you can share your ZIP code (for best weather station data) and some January-February time frame gas bills with the exact METER READING (not billing) dates, and amounts, and the input & output numbers of the older equipment.
Seems there is not sufficient enforcement of CA Title 24, which IIRC requires Manual-J heating & cooling load calculations even on replacement equipment, and may even require duct leakage testing & remediation (though that might only be for new construction.)
@ Dana
"A typical 2x4 framed house with clear glass double-panes (or storms over single panes) with 1550' of all-above-grade conditioned space would typically come in around 22-28,000 BTU/hr @ 0F outdoors, 68F indoors. A 2x6/R20 house with U0.35 (no longer meets code performance north of US climate zone 4) would come in 19-24,000 BTU/hr."
My house is about 700 sq/ft 2x4 cellulose insulation double pane vinyl windows, single floor with 2/3 basement, 1/3 crawlspace -2F 99% temp (near Toronto Ontario) and is rated in Hot2000 at 45.5k peak load. Now thats with the massive air leakage but even if i reduce it to 0.1ACH50 i get 32.7K so i don't understand where 22-28k is coming from at double the size.
Alan B: Measuring fuel use against heating degree days in reasonably tight houses (not super-tight), that's where it tends to come out, and it's usually within 15% of an aggressive Manual-J calc. There may be something amiss (assumptions not aggressive enough) in your Hot2000 inputs other than mere air leakage.
I've not used that tool more than a couple of times, but IIRC that Hot2000 losses from uninsulated basement slabs seemed high, but not enough to double the whole-house load.
The shape of the house also makes difference in terrms of exterior surface to floor area ratio. My house has a ridiculous number of corners and three dormers-not exactly an efficient shape, and 10-12 years ago I was surprised that it was performing better than my initial heat load calculations indicated, I've since learned to be much more aggressive about the R-value & infiltration assumptions, as well as more careful accounting of the 24/7 plug loads.
I wouldn't be surprised if a bay area antique was air leaky enough that it fell above the high side of my estimates, but not 2x, and CLEARLY not 3-4x! (Which is what it would take to make a 60K furnace right-sized.)
@ Dana
I once tried to recreate my house in BeOPT, it would not get the roof right and i gave up (it insisted one half was perpendicular to the other half instead of connected)
Believe me, i would love to get under 25K for my house, i could eliminate the high fire on my 40K furnace, right now it kicks in for under a minute @ 5F (as well as reduced load would cut the heating bill quite a bit)
I do agree that a Bay area antique can't need 60K which is what many installers were trying to sell me (old furnace was 75K/60k out).
Alan- since you have a heating history to work with, have you run a fuel-use based heat load calculation on your place? (I'd be surprised if it was even as high as 20,000 BTU/hr @ -2F, and it could easily be under 15K if you've sealed & insulated the basement & crawl space.)
A load/floor area ratio will typically be higher for a smaller house than for a larger house due to the higher ratio of exterior-surfaced/floor-area, but ( 45,500 / 700' =) 64 BTU/hr per square foot @ -2F just isn't anywhere near credible number even for a 700' house.
I'd be surprised if locking out the high fire on your 40K 2 stage would put you in danger of losing ground at the 25K low fire. At 95% efficiency that's 23,750 BTU/hr or ( / 700'=) 34 BTU/hr per square foot. The crusty old-schoolers in my area (outside design temps = 0F ito +5F) use 35 BTU/ft^2 -hr for older uninsulated or barely houses for furnace sizing, and still manage to oversize by 2x or more.
John Melichar's Goodman GMVC960603BN has a low fire output of 40,320 BTU/hr (which is more than the heat load at my house) and the whole-house load is probably on the order of ~10,000 BTU/hr @ 39F. AND it's broken up into two zones. In all likelihood the furnace is about ~10x oversized for the 500' zone even at low fire, and 5-6x oversized for the 1000' zone at low fire.
If it a complete DO-OVER were even remotely economically viable, it would probably make more sense to do it with a condensing tank hot water heater and a tiny one ton hydronic coil air handler like a Firstco 4CW, which delivers ~13,700 BTU/hr with an entering water temp of 120F @ 2 gallons per minute pumping rate and 19,200 BTU/hr at 140F EWT @ 2 gpm according to the charts on p.3 of the spec sheets:
https://www.firstco.com/getattachment/Products/Commercial/Wall-Closet/CW/CW(X)0915.pdf
That would deliver long comfortable low-cfm flow cycles. The max flow on the Firsco 4CW is 400cfm, which is lower than the lowest heating cfm setting on the Goodman GMVC960603BN (600cfm, set to minus 10%.)
If air conditioning is a "must have", doing the heating & cooling with mini-splits might make more sense. But cooling loads are low enough on that side of the bay that most houses that size do just fine with a half-ton window-shaker for cooling.
Prepare to be surprised ;)
Jan 7 - Feb 8 2016
-2F design temp
309.841 m3 used - 15.5 gas stove/water = 294.341
Fahrenheit-based heating degree days for a base temperature of 65F = 1151.3
Fahrenheit-based heating degree days for a base temperature of 60F = 991.3
High input 40,000 output 39000 = 97.5% Low input 26000 output 25000 = 96.154% (low stage used for calculations)
283.02m3 of heat = 101.947 therms - 10,194,700 BTU 10.194700 million btu
10.194700/1151.3 = 8854.9466 @ 65F / 24 hours a day = 368.956
10.194700/991.3 = 10284.1723 @ 60F / 24 hours a day = 428.507
65F base = 67F delta T x 368.956 = 24720.059
60F base = 62F delta T x 428.507 = 26567.445
24720.059 * 1.4 = 34608 (including hot water tank 309.841/294.341 ) = 36431
26567.445 * 1.4 = 37194 (including water tank 309.841/294.341 ) = 39153
The basement and crawlspace are not insulated, they are made of brick and i am concerned that rigid foam or spray foam could cause damage to the brick foundation over time from freeze/thaw cycles, especially with some of it being above grade and cutting off the heat flow.
Also i have gaps from basement/crawlspace to attic (balloon framing) that i can't address at present till the crawlspace is dug out and encapsulated, the house is made from true 2x4, i keep the temp at about 70F and the energy audit found 14.85ACH50.
A couple years back we had some -15F weather and it would kick into high fire (39k output) for a while and would shut down even if only for a few minutes so its definitely not undersized for the house
I do not understand this 60F and 65F balance point you speak of but i tried to follow your instructions to the letter
Hot 2000 - ANNUAL SPACE HEATING SUMMARY
Design Heat Loss at -2.20 °F (4.62 BTU/hr / Ft3): 46642.80 BTU/hr
Gross Space Heat Loss: 105.51 Mil.BTU
I will add another data point.
My house is 2400 ft^2 above grade with another 1800 ft^2 of conditioned basement. The 2x6 walls have R-10 of exterior foam. My manual J predicted a heat load of 34,794 btu/hr at an inside temperature of 70 and outside temperature of 7. My I=B=R spreadsheet, based on Marc Rosenbaum's template, predicted 32,869 for the same conditions. I have fuel use data for October 30-December 2, so I just calculated the load using the fuel use method. Using the 60 degree balance point, I get 23,612 btu/hr. Using the 65 degree balance point, I get 18,234 btu/hr.
We have a forced air gas furnace rated at 39K btu/hr on high stage and 25K btu/hr on low stage. (I had serious pushback from the HVAC contractor on that.) On nights with an overnight temperature near the design temperature of 7, the furnace doesn't quite keep up. The furnace runs continuously and the thermostat temp drops back a degree or two by morning. We get a lot of solar gain. Once the sun comes up, the upstairs zones may go all day without calling for heat even though the outside temperature only increases about 10 degrees. (The basement zone still calls.)
Alan B: "A couple years back we had some -15F weather and it would kick into high fire (39k output) for a while and would shut down even if only for a few minutes so its definitely not undersized for the house"
Note, -15F is fully 13F below your 99% outside temperature bin, and it only needed to run at high fire for "...only a few minutes...", which would indicate that your heat load at -15F if it's the load at that temp si above the low-fire output, it its is barely above the low-fire output of the furnace, Most 2 stage furnaces only step up to high fire if the thermostat isn't satisfied within a pre-determined amount of time. If your furnace operates that way, it doesn't necessarily mean the indoor temperature would actually lose ground with the high-fire locked out, only that it takes longer to satisfy the thermostat if the high fire is locked out. If right sized for the load at -2F it would be needing to kick into high-fire sometimes even when it's above -2F.
Reid: It's rare than an I=B=R would come in lower than a Manual-J, which probably indicates the assumptions in the Manual-J aren't aggressive enough.
Clearly we overbought and we will hire a consultant to look over the system and determine what size furnace we should buy.
I will also ask whether we should consider 2 systems rather than zoning.
I'll also ask about the cost, but it's a price we pay for expecting this general contractor to know about this and not doing enough research ... but how often does a person buy a furnace?
Given we have to live with the system for at least 10 years and hopefully more (our old furnace had been in the home since 1979) I want to do it right.
btw In December of 2015 our therms were the highest at 95 and the lowest was 21 therms in September; we have a gas water heater and stove.
Between what exact meter reading dates was that 95 therms used? And, what is your zip code (for the nearest weather station, most accurate data.)?
Unless you need the ducts for air conditioning it might be easier/better to go with a hydronic (hot water) solution with radiators or baseboards, running it off the water heater. If it's going to be ducted hot air, a hydro-air handler running off the water heater is probably better than any furnace. Unlike furnaces, there isn't a minimum air flow restrictions to worry about.
I found another bill from Dec 24 2015 - Jan 25, 2016 that shows the therms at 122, two days being the highest at 7.21.
Zip is 94132.
I had considered radiators or even radiant heat in the floors, but was vetoed and the 500 sq ft we remodeled is framed, drywalled and painted; we've also got a tankless water heater now if that factors in.
I appreciate getting a better understanding from all of you.
@Dana, you didn't read it the way i meant it, it would shut down for a few minutes here and there, so say its -15F outside, in an hour its shut down maybe 2-3 times for a couple minutes, so its running maybe 50-55 minutes out of 60. Granted some of that is low fire but the total heat load is well above 25k otherwise the 1.56 dial up factor (25 to 39k) would not be necessary to satisfy the thermostat for only a short few minutes. So the load is under 40K, but its probably 30-35k, hence i can't lock out the high fire.
So i still don't see where 22-28k @ 1500sq ft at 0F is coming from
Don't forget my bill use done with your formula is about 25K and closer to 40k at 1.4x.
Dana, you were surprised that my I=B=R was lower than my manual J. Both are based on the same temperature assumptions. I provided the same information about R-values and U-factors to the people that did the manual J as I used in my spreadsheet. Why would a properly done manual J be expected to come in lower? Does manual J give credit for some internal gains?
Alan, your furnace probably runs 10-12 minutes on low and then switches to high. Then, it starts over on low on the next heat call. If there are three cycles in an hour, then it is still spending 30-36 minutes per hour on low.
@Reid: "Does manual J give credit for some internal gains?"
Yes- Manual-J accounts for 24/7 plug loads such as refrigerators & DVRS, the numer of occupants, standby losses on hot water heaters etc.
@ Alan B: "So i still don't see where 22-28k @ 1500sq ft at 0F is coming from
Don't forget my bill use done with your formula is about 25K and closer to 40k at 1.4x."
22-28K @ 0F is where most (but not all ) houses that size seem to come in. I haven't tried to keep hard statistics on it, but that's my gut feel for the first standard deviation. Yours seems well outside of where a reasonably tight reasonably insulated house would be, but with no basement or crawlspace insulation and air leakage at 14 ACH/50 it would be outside outside of that description, whereas John Melichar's house is most likely both much tighter and better insulated than that, having recently been upgraded on insulation.
BTW: "I do not understand this 60F and 65F balance point you speak of..." Houses have internal heat sources (and also trap solar gains.) So there is an outdoor temp well below the indoor temperature at which the heat accumulation from the internal gains balance with the heat losses due to the temperature difference. Above that outdoor temp there is usually a sensible cooling load. The reason the heating degree day base temperature has traditionally been set to 65F/18C is because that is the typical balance point for homes back in the 1940s & 1950s. But as homes have gotten tighter and better insulated the balance point between heating & cooling has dropped. Most newer code-min houses have balance points around 60F. Reid's 2x6 +R10 house probably has a balance point below 60F. With enough data that balance point temperature can also be determined empirically. (Degreedays.net has an app for that: http://www.degreedays.net/regression-analysis )
Reid's house has also sounds like it has higher than typical solar gains, which introduces another source of error in a fuel-use load calculation, but when hot water heating uses the same fuel it counters the solar gain error, sometimes over-compensating.. For John Melichar's house I would anticipate the hot water use error to be larger than the solar gain error.
@John Melichar: Exact meter reading dates are necessary for correlating that use to the real weather data, and in locations as temperate as yours even a couple of days error in the meter reading dates can add up to a double-digit percentage error.
@Dana
That is really interesting, it makes a lot of sense now from the explanation.
Thanks :)
I will try the calculation again when I have fuel use data for December and January. Right now, I only have October and November. I expect that the result will be less sensitive to the balance point assumption when the weather is closer to the design temperature. I will try out the regression tool at that time.
I am not sure how to treat my hot water heating impact on this analysis. I have a heat pump hot water heater. It doesn't directly account for any natural gas use. However, at a COP of 3, 2/3 of the water heating load is indirectly satisfied by natural gas via the furnace. When the hot water heater is operating, it draws about 4k btu/hr from the house which is not inconsequential relative to the envelope heat loss. However, most of the time that the water heater is recovering occurs during the time the morning sun is giving me solar gain.
The stove and dryer are electric.
I do not know about Goodman but the Lennox 95% 2 stage I have has a setting for minutes on low fire. It came from the factory set at 10 minutes, I moved the switch on the control board to 15 minutes.
(Finally) taking a look at John Melichar's fuel use between 24 December 2015 and 25 January 2016, if one takes the average of the base 65F totals from 24 Dec-24 Jan (442.5 HDD ) and 25 Dec-25 Jan (435.4HDD) it comes out to 439 HDD. Without subtracting anything from the 122 therms for other used the implied load @ 39F is 24,088 BTU/hr if it was an 80% efficiency furnace, or 28,604 BTU/hr if 95% efficiency.
If one allows 20 therms for other uses (likely on the low side, given his all-time low monthly usage of 21 therms) the implied loads are 20,139 BTU/hr if 80%, 23,915 BTU/hr if 95%.
Taking the largest of the hot-water use adjusted number and using the ASHRAE recommended 1.4x multiplier the ideal furnace would have an output of no more than ~33,481 BTU/hr. If it was an 80% furnace it should be no more than 28,195 BTU/hr.
See the attached spreadsheet calculations.
Those might be credible numbers for the "before upgrades" picture, but would be extremely high for the "after windows, air sealing & insulation" picture. Last year I did a similar web-forum consult on a 2x4 framed 1500' house + full basement house in Portland Maine for boiler sizing purposes. While that house reasonably tight, it had no foundation insulation, and the implied load at +6F was about 33K, and with modest improvements would be under 30K, and that's at more t than TWICE the temperature difference. After installing the new boiler and dialing in the outdoor reset curve, and some spot air sealing the implied load is now about 25,000 BTU/hr.
Unless the ducts are outside of conditioned space and leaking, driving air infiltration, I would hazard that the "after" picture of John Melichar's house is really under 15,000 BTU/hr, but that can probably be verified by this winter's gas billing.
One can probably use the existing brand-new tankless and an ECM drive bronze-impeller pump with a programmble gallons-per-minute, with a 1-ton ECM drive hydro-air handler to deliver a more reasonably sized heating system using the same ducts. With some tankless vendors that would violate the warranty, but at these low loads it's not going to burn it out if you keep the flow under 3 gpm with at least a 15F delta-T. These are not tough systems to design, nor are they very expensive. At 2gpm and a 20F drop across the air handler coil (say 130F out, 110F back) it would be delivering 20,000 BTU/hr to the house. At 3gpm and a 15F delta it's 22,500 BTU/hr. With some tweaking one can find the right compromise of flow & water temp & air handler speed to deliver reasonable recovery rates from setback (say, less than an hour), and reasonable exit temps at the registers, etc. As long as the water returning from the air handler coil is under 120F any condensing tankless will deliver efficiency north of 90%, some north of 95%.