Knight vs. Weil-McClain
I have an unusual situation: Three buildings. Two are 2,000 sq. ft. One, on two floors is 4,500 sq. feet. These are external building dimensions. Radiant heat everywhere. Propane. The original house is 2,000 sq ft. and is currently heating by a Bradford-White Combi Hotwater Heater. We are replacing it. We are about to put in a boiler in the larger building, which will supply heat for the radiant floor and for domestic use for all of the buildings. I have gotten it down to deciding on either a Weil-McClain or a Lochinvar Knight. I would like advice as to which is best and most trouble-free. Also, estimate of size needed? I estimate about 200,000 btu. Is this correct? The buildings are well insulated. Many thanks.
There is no way to tell the required size of the boiler based on what you gave...while there are some WAGs based on square footage, they are usually way off. You need to know the exact amount of insulation, window location and sizes, orientation, and the needed design temp along with the heating degree day info for your specific area. You'll save money and grief if you pay someone to do this but, not everyone is capable of doing it, or willing to do it right (they'll just give you a WAG). You can figure it out yourself if you have fuel use history along with the HDD info, which you can download for your area. Each pound of propane provides a known (or pretty good estimate) amount of heat. That coupled with the efficiency of the existing boiler and you can see how much heat is currently being used to heat the place. That sized boiler could be as much as 3x bigger than needed, and if so, would be lousy in efficiency and longevity. The only way to tell is to run some numbers, and you've not given enough info to do it.
Both manufacturers make good products- the quality of the system design and the installer backing it up is FAR more important than which manufacturer. Ask your local/regional distributor if they can recommend heating contractor. The distributor is in a unique position to determine who is installing dozens of them with minimal factory support, and who is constantly screwing it up and bugging them for info that's already in the manual.
What Jim said about heat loads. Terms like "well insulated" are meaningless, as well as "BTU per foot" methods of heat load calculation. Fuel use against heating degree-days can put an upper limit on it, but a room-by-room heat load calculation is the tried & true method, as long as there are no thumbs on the scale like design temps that are well below the 99% outside design condition, or mis-stated U-factors and contruction types.
If I'm reading the description correctly, you have 8500 square feet of conditioned space you're trying to heat with this boiler? If yes, 200,000BTU/hr is likely to be well over your actual heat load. Most so-so insulated buildings in PA are going to work out to about 15BTU per square foot of conditioned space, give or take 5, depending on how much window area (and type), and how much air leakage you have. Most "well insulated" and fairly air-tight houses will come in under 12BTU/foot, sometimes under 10BTU/ft. Your estimated 200,000BTU/hr is knocking on 25BTU/ft, which is would be a typical the ratio in places like Fairbanks AK, with design temps WAY below yours. (With a zip code we could zoom in a bit closer but your 99% design temp is probably around +10-12F, which is ~40-50F warmer than design temps in central AK.) It wouldn't surprise me in the least if a heat load calculated on fuel use or construction type U-factors pointed toward an 75,000-100,000BTU boiler and I'd be quite surprised if it needed a 200,000BTU boiler.
All good heating system designs start with the heat load calculation. Even if you have to pay an energy nerd to run a careful "Manual-J" , which may take several hours, if it's three distinct buildings with differing construction types, it'll probably pay for itself on day-1 in reduced boiler size, and every year thereafter on efficiency and reduced boiler maintenance.
If you oversize a boiler you risk short-cycling it- the smallest zone needs to be able to take the full output (or full minimum-fire output at the lowest operating temp, if it's a modulating condensing boiler), without short cycling. The absolute smallest boiler that can meet the heat load at the 99% condition is what you're after, since it'll have the longest and most efficient burns, and the fewest thermal cycles. If all of the radiant floor is massive concrete slabs it's less of an issue than if it's tubing under a wooden subfloor or something, but whomever is specifying the boiler needs to do at LEAST the napkin-math on the whole system design, and it's likely that the ideal operating temperatures of the different zones will differ, and it's probably worth the extra system design & plumbing to run them all at the lowest possible temperature (=highest efficiency).
If you know your window U-factors and total window square footage, your wall construction type & insulation values, and your attic/roof areas & R values it's pretty straightforward to come up with reasonably accurate heat loads calc based on construction type using standard spread sheet tools, taking a WAG at or inserting fudge-factors for air infilration rates,etc.
Thanks, Dana and Jim, for your response. The fellow who designed the radiant heat tubing (and was going to do the boiler installation) did heat calcs a long time ago. (We have been building for 12.5 + years, as money allows, and the calcs were done about 10 years ago). He came up with 200,000 btu. or slightly less. I will hire someone to do new heat calcs. We're in zip code 18942. Some of the rooms, a large storage room, a carpentry shop, and a guest suite, which will get rarely used, will be kept at no more than 50 degrees (except when the guest area is in use). A greenhouse will be kept at 45 degrees. Other spaces (there are 11 or 12 zones) will be kept at varying temperatures, depending on their function, but no more than 65 degrees. Insulation in the ceilings to the roof in two of the buildings is 3" spray urethane foam, in the other it is fiberglass. Walls in the fibreglass roof building are 2" spray urethane foam (once had spray foam on the roof, but due to a design flaw had to replace the roof 24 years ago and at that time could not afford to replace the foam). The walls on the other two-story buildings are 3" urethane foam boards. Windows are all fixed thermopane. (Custom vents underneath for air flow in warm weather.) All of these buildings are a big dream for us--we are photographers as artists--scraping by month to month, but whatever we do, we only want to do right. Which is why it has taken so long. We work at home, which is why we need all of this space--far more than needed just to live in--2,000 sq. ft. is plenty for that. We actually live very simply. Again, many thanks.
Looking at Weatherspark.com data, Ottsville's low temps track those of Allentown's so using Allentown's 99% outside design temp of +10F would be the right number in any load calc. (Yes, it get's down to 0F sometimes, but that's less than 1% of the time, and the thermal mass of the buildings carry you through those relatively brief dips if if the heat load calc and boiler had ZERO margin.)
The greenhouse may have the largest heat load at +10F, since it has the highest U-factor.
A 3" closed cell 2lb density polyurethane foam is about R18. At 3" of open cell half-pound density polyurethane foam is about R11. The air-tightness is pretty good, but these are below current code-min R values for residential buildings in PA.
Rigid board insulation is rarely made of polyurethane. Polyisocyanurate (iso) is an off-white slightly yellowish look and always has either a fiber/papery or bright foil facer. Extruded polystyrene (XPS) is uniform in color throughout with no facers, and it's usually pink,blue, or green from the most common vendors. Expanded polystrene bead (EPS) is usually white, with obvious bead/grainular structure boundaries (it's the same stuff as cheap coolers and coffee cups, but higher density), and sometimes comes with plastic or foil facers. Iso is ~R6/inch of thickness, XPS R5/inch, and EPS is ~R4/inch. At 3" only iso would meet current code min for wall R-values.
Thermopane varies in U-factor from ~U0.22 (low-E coatings on at least 3 of the window surfaces) which is about R4, to about U0.6 (clear glass, no coatings) which is less than R2, so it really matters what you have. If it's pre 1985 assume it's no better than U0.6.
U-factor is BTU per square foot per degree-F difference. The U-factor of an insulated wood framed building depends on both the insulation and the thermal bridging of the framing. A studwall with 3" of R11-R18 of foam between the studs is between R9-R11 after the thermal bridging of the wood is factored in, so for rough purposes call it R10. The U-factor of the wall is 1/R, or 0.1 BTU per square foot per degree-F difference. If it's rigid foam on the EXTERIOR of the sheathing (or roof deck) that is continuous, with no bridging rafters are studs, the R-value and U-factor of the assembly is the same as that of the foam. So 3" of continuous iso (R18) would have a U-factor of 1/18= 0.056 BTU per square foot per degree-F.
So, start measuring the square feet of each assembly & construction type, and add it all up. Subtract out the window areas from your wall dimensions, and add up the window areas separately, using a U-factor of 0.6.
eg: Say your green house has 200 square feet of U0.1 wall, and 400 square feet of U0.6 window, no roof. You're keeping it at 45F, the design temp is 10F, so the delta-T is (45F -10F =) 35F.
The heat loss from the walls is then:
200ft x U0.1 x 35F= 700 BTU/hr
The heat loss from the glazing is:
400ft x U0.6 x 35F= 8400 BTU/hr
The total heat load for the green house is then something like (700 + 8400 =) 9100 BTU/hr.
For solid wood exterior doors, use U0.5 for the door U-factor. For panelized, use U0.75.
For the section of roof that has the fiberglass insulation between rafters, use U= 1.3/(fiberglass thickness x 3.2)
eg: If it's 5.5" thick batts the U-factor is about 1.3/(5.5 x 3.2)= 0.074.
Run the numbers on the whole place, using your standard thermostat settings (not the "guests only" values) for the delta-Ts. Add it all up, and you'll have a pretty good idea what the lower-bound is. The true heat load will be higher than that due to air leakage and possible gaps in the insulation, but it won't be 2x, it won't even be 1.5x. If the place is pretty air tight and draft free you can just run with it, but if the buildings have known drafts, add maybe 15-25% and you should be good. Don't upsize from your calculated number by more than 50% or you risk running at below the rated AFUE on the boiler.
If you hire somebody to do the calc using a standard heat load software package, be sure to look over their shoulder to get the right indoor and outdoor design temps. If they enter 72F indoor temp and 2F outdoor temp and you're actually running 45F/10F, their calculation would be twice your actual load. They need to understand very clearly the temperatures you intend to operate it.
Most of the time your temps are well above the outside design temp, and raising the guest quarters to 65F or higher won't be a problem. Design the system to run efficiently for the average load and still meets your min-indoor temp numbers at the 99% outside design temp. Don't buy an oversized boiler just to meet the <1% condition that have house guests on the very coldest day of the year. Odds are it would keep up anyway even at 15% upsizing from your spreadsheet number, and if it doesn't, a $50 electric radiator type 1500W space heater would usually make up the difference. The fact that much of the space is maintained at 45-50F takes a HUGE chunk out of what would be a typical residential heat load.
BTW: Many 1.5 ton mini-split air source heat pumps can put out about 22,000BTU/hr at +10F with a 70F room temp (and much more at lower room temps.) The average efficiency is quite good- your average coefficient of performance would be over 3.0 if you're only keeping the place 50F, or even 60F. And the cost of the power used would be less than half what it would take at recent propane & electricity pricing. The installed cost of a high-efficiency 1.5 ton mini-split is ~$4-4.5K, but it's also possible to do a DIY install for less than $3K. A 2-ton mini-split is only about $500 more, and could put out 30,000BTU/hr @ +10F. These things modulate output with load and run most efficiently in a "set and forget" rather than a setback strategy. It may be the "right" way to go for some of your spaces, since the reduced operating costs might pay for the thing in short years compared to propane pricing. Think about it, after running the room by room, or building-by-building heat load numbers. Since some mini-splits don't have settings below 60F, run a separate sheet using 60F as the indoor design temp. Greenhouse growers use these things a lot in lieu of propane burners, since it's often just 1/3 the operating cost.
Thanks so much, Dana, for the time and effort you have just put into responding to me. It will take me a while to go over all of this, but there is no question that I will do so. More details: It is amazing to me how much I left out and also how much I do not know. You are so kindly giving me quite an education. All of the buildings are made out of either 12" or 8" cinder block and are insulated on the outside. The foam board is the stuff that looks like spray urethane foam. It is dense and not at all crumbly. It has black tar-paper covering on both sides. In the larger building with two stories, the west and south sides of the bottom floor are under ground. Views here are to the north. Unfortunate orientation, but the site dictated that. The buildings were designed by the Design/Build team "Jersey Devil" to be passive solar so that helps a lot in winter. "Jersey Devil" has two books of their home designs out there and they are fanatic about proper engineering.
Other: Since I will definitely be going smaller, and less expensive, than the 200,000 btu boiler I thought I needed, would a Buderus be the preferred choice? I understand it is the installation ans sizing that matters most, but I have to ask. I have heard the Buderus referred to as the Gold Standard.
I sure wish I could hire you to do all of this--the calcs, the install. I'm sure many on this forum feel the same way. Thanks again.
One other spec. All exterior doors are 3" solid wood.
For estimation purposes, use U=0.35.
Originally Posted by michaelasmith
The asphalted paper facers are commonly used on 2lb density roofing iso, figure R5.5-R6 per inch. (And U=1/R) But there were other similar products 20 years ago that were also ~R6/inch. With cinder-block + 3" foam + some sort of finish siding & interior, use U=.057. If there's no interior finish wall and something thin like vinyl or metal siding, try U=0.060. (There's probably more error in the guesswork about the insulation type & R value than the difference any siding would make.)
Buderus also makes good boilers and has good regional support, and people sometimes pay a premium for them. But a talented hack designer can bring ANYTHING to it's knees, eh? ;-)
If the radiation is already installed the manual j calculation is helpful but if it doesn't match the installed equipment you will have a problem matching boiler to load
Actually, all the information Dana listed (I wish I could type that fast) is useful to a radiant floor heating system designer and can certainly be plugged into a dedicated radiant floor heating software program based on Manual 'J'. The sum of the information properly input into a room-by-room, or at least a building-by-building heat load program will allow for very accurate sizing. As Dana properly points out; it is not about the equipment brand so much as the size.
More important to me is the specific type of radiation and the room loads all of which will determine the what the condensing boilers maximum and minimum output should be.
We use Wrightsoft with an Uponor module among others with 13 different condensing boilers to date. We install and service many condensing boilers featuring aluminum heat exchangers including Buderus, Bosch, SlantFin and Weil McClain but we also know how to use a pH meter and combustion analyser. Before you start "hacking" get a professional to back you up.
Thanks to everyone for your input. I have referred our radiant heat guy to this thread. He did heat calks,but I do not know on what they were based, I will find out early next year. And will report.