Actual heat loss vs. calculated heat loss....

STEVEusaPA

It occurs to me from reading, and my own experience, that even the best heat loss programs tend to oversize equipment from 10% to (some people claim) 50%.

Was just wondering if there are any studies, or if anyone ever studied the following:

Trying to do an actual heat loss at the home.  I realize there are a ton of variables, but thats where computers could calculate away.

Basically, measure the temperature of a central room.  Run the burner for 1 hour, re-measure the temp and calculate how many btu's it took to raise the temperature that many degrees.  Then with the heat off for the next hour, measure temp, and how much temperature the house loss in the next hour.

Like I said, the variables would come into play, outdoor temp, wind, sun etc.

But my thinking is, you're going to spend a couple hours measuring everything, and putting it into a heat loss program anyway, only to get an educated guess.  Maybe there's another way, or a better way then what I'm thinking.

Just random thoughts from someone with too much time on my hands via the warm winter.

Thanks,

Steve

Jean-David Beyer

Run the burner for 1 hour, re-measure the temp and calculate how many btu's it took to raise the temperature that many degrees.

I do not think that would work for my house. If I raise the temperature of my thermostat, it takes 4 to 8 hours to change the room temperature, and about 24 hours for the system to really stabilize. House is radiant slab at grade downstairs. By the next day, the outdoor temperature may have changed 10F to 20F because of Murphy's law, so re-running the experiment again would prove not much.



Similarly, if I lower the temperature of the thermostat, it takes many many hours for the slab to cool down.

PeterNH

Electric

Hello Steve,

In my, highly insulated, 500sq. ft/3200 cu. ft.. model railroad hobby room there are lots of lights and electronics.

More than once i have forgot to turn it all off at night.

The next morning it would almost always be warmer than the thermostat setting.

Having a Kill-a-Watt meter, i set the light to equal aprx. 1kw/ 3415 btu's for a week, with the heat off.  It could maintain the temp down to about 10 degrees with minimal wind and the garage temperature below at no less than 46*.

The heat loss calc i have done,Siggy's HDS., gives me aprox. 5130 btu's at -3*.

The empical evidence from above calculates to about 4525, at -3.

Overcalculated by about 13%.

Not bad at all.



Peter





The radiation is 15 feet of fin tube and has always been able to keep up



even  in -5 to -10 nights

STEVEusaPA

werent you suppose....

to post a pic of the model RR room?

PeterNH

I did

 it was in the other thread.



I promise to post some new pics next week.



Peter

Jamie Hall

In principle...

you're right on track.  The devil, as they say, is in the details!  Jean-David pointed out one set of problems -- the thermal inertia (heat capacity) of the system.  The way to get around that, of course, is to measure the amount of heat it takes to keep the house at a constant temperature which nicely gets rid of that variable!  I think one could get rather close if one did that -- on a day when everything outside was also pretty constant; say a grey overcast day with no wind and a pretty steady outside temperature (we get them now and then up here in New England).  So... I'm inspired to try it and see what happens!



The problems -- at least in this ark which I manage -- come in from two sources: one is the sun; despite the fact that no one was thinking about solar gain when this place was built, it picks up a surprising amount of heat from the sun.  Nice of it.  The other probem is the wind -- even if this weren't a draughty old place, the wind has a huge effect on heat loss; in a draughty place like this, though, the wind makes a huge difference.



Have to think about this.  Sounds like an interesting sort of project!  I have much of the data to do it as an average already... hmm.  Maybe when I have too much time on my hands I'll give it a go!

STEVEusaPA

Jamie...

thanks for your response.  If you do give it a go, I would definately be interested in what you come up with.  I also figure someone, somewhere has tried/comtemplated this, probably a DeadMan from long ago....

The Steam Whisperer (Formerly Boilerpro)

How I calculate

While not as perfect as running scientific studes, just looking at your historical fuel usage, estimating the efficiency of your heating equipment and looking at degree day data will get you alot closer than all the heat load calcs you can do. 

Besides solar gains, and other internal gains, you also can have ground contact tempering on high mass stuctures along with the thermal flywheel effect.  I have never seen any load calculations take ground contact into account, however, advanced cooling load calcs do take thermal mass into account and length of occcupancy.    That's why for many church's that I've run cooling load calcs, a crowd of 150 people on a sunday morning in a high mass building doesn't need more than 4 to 6 tons of cooling, despite all the other contractors saying it would need 12 tons. 

Jamie Hall

Boilerpro's approach

is pretty much what I have data for.



I have the average number of gallons of oil used per day per degree-day for this ark: 0.41.  I have the total exterior area: 8,450 square feet.  Now with a very minor amount of arithmetic, one can move the numbers around and weave a magic spell or two and come up with an apparent R factor for the entire envelope of 3.5; corrected for boiler efficiency which Charles tells me is about 85%, the actual envelope R comes out to 4.2.  Which, I admit, sounds horrible in these days of super insulation and all that, but keep in mind that the building walls are not insulated (some of the roof is); the windows (and there are a lot of them!) are all double hung with storms (it's a National Register site, and there are definite limits on what can and cannot be done in terms of restoration/renovation -- and budget); infiltration is very significant -- and it sits on top of a hill exposed to all the winds that blow.



for what it's worth...

STEVEusaPA

thanks for the input again Jamie.

So after your figuring, do you feel that:

1.  The boiler is matched to the connected load?

2. The total EDR initially spec'd for the house was correct? Or do your numbers reflect too much radiation was iniitially installed.

I do like boiler pro's idea.  I just need a real world example to put it into perspective.  A heating season's data will ultimately give me how many btu's were needed for the season/dday's for the season, equaling how manybtu's/ddays/per day.

This tells me what the old equipment did at it's measured efficiency.  And it would tell me fuel savings based on the same size equipment, for a higher efficient unit, which could be used to reduce the new equipment based on efficiency alone.

How would I make the leap to convince myself how much smaller I could go?

Thanks for the input.  Keep in mind, it has to get through my thick skull

Jamie Hall

Oddly...

I do have the data for the original install -- in 1930!  The install was meant for a 70 degree differential, and it would seem -- from the way the building responds on a still day with slightly less than that differential that it was overdesigned by about 20% -- for still air.  It is most assuredly not overdesigned when one takes into account the wind.  Then one realises that it was designed assuming that the gravity hot air furnace which was installed in 1893 was assumed to be operable, as the section of the building which was served by that furnace is not adequately heated if there is any wind at all.  That furnace has died -- in fact, died about 40 years ago -- although the body is still lurking in the basement (and is real scary to look at). (Incidentally, the present boiler is about the same BTUh output rating as the original -- a truly impressive H.B. Smith -- but burns about half as much oil).



The boiler -- thanks to a combination of my numbers and work on the system but particularly to Charles Garrity's careful work on the boiler install and tuning -- is almost perfectly matched to the availale radiation.  It does cycle on pressure (6 ounce cutoff / 0.5 ounce cutin) on very long runs; in fact the chosen night setback (3 degrees F) is based on just reaching the cutin on a still day, which takes about 45 minutes or so.



On a windy day, all bets -- and a lot of BTUs -- are literally out the windows...



On a more philosophical note, it is my impression that carefully calculating the building envelope effective R factor, and sizing the radiation to it, works -- with several cautions.  First, one must use the overall effective R factor.  Just because one has say R 19 walls does not mean an R 19 building!  Second, one must account for the probability of infiltration, and extrapolate that probability down the road some: just because a building is very tight when it is built does not mean it will stay that way for the life of the installed radiation.  Third, one must size the radiation in a space to the specific requirements of that space in terms of wall and window area and exposure.  Simply applying some factor to the square footage of the building (or even the footage of the exposed wall) to come up with a heating load is either going to produce some cold areas -- or result in an overdesigned system (if I were obliged for some reason to do that myself, I'd overdesign by a factor of at least 2, depending on how many lawyers were related to the client).

Robert_25

"It depends"....on a lot of things!

Steve, I think a total heat-loss calculation tends to be more accurate on new construction...because (if everyone did their job) you should know how much insulation is in the walls, etc. 



For an old house like mine, a "best guess" heat loss calculation is needed to size the radiation per room and such, but historical fuel consumption and/or boiler run time seems to be better at telling 'the real story' on the heatloss of the entire structure.



When I first moved into this house I did a heatloss calculation with the Slant Fin software.  I had to guess at the insulation values and came out to 120,000 btu's per hour at -15F...no wonder the 250k btu oil boiler tended to short cycle!  I installed a Beckett Heat Manager to help with the short cycling (it did) and an hour meter, so I could keep track of the burner run time.  I soon learned that the boiler rarely ran more than 5 or 6 hours per day.



For a while I kept a log book next to the boiler and tracked the burner run time vs the high & low temperatures of that day.  I don't have the numbers right in front of me, but from memory I think the coldest day I had in the log was -22F in the morning with a "high" of -5F in the afternoon...even with those temperatures the btu INPUT to the boiler didn't even average out to more than 100,000 btu's per hour.  I now know that the wall cavities are filled with rock wool and/or cellulose, so the old house is better insulated than I originally thought when I did the heat loss calculation.



One winter of buying $3+ fuel oil was enough, so I installed an anthracite coal stoker boiler, and I have it set at approximately 100,000 btu's per hour.  To-date it has never run more than 15 hours out of 24, and that particular day wasn't the coldest of the season...it was cold and windy.

icesailor

Reality of Life:

Back in my earlier life, I wondered and worried about this stuff. I lived and worked in a place that had extreme weather conditions. I decided to figure out how to calculate heat. So I could be more efficient. What I learned is that even though they may tell you that they have a "design day" in mind, there will come a time that that figure isn't low enough. One day, I remember such a day. It was 5 lebelow and blowing 35 and gusting to 70 degrees. That isn't in the tables. I remember trying to turn the water off in a house that the tenants tried to save on heat so they turned down the thermostat so the heat pipes in the wall froze and the water pipes above froze and broke. There was water in the meter pit and while I tried to shut the meter off, in a prone position in the ice and snow, ice made up on my arm and a cop stopped to see if I was a dead body. When you have those conditions, you really don't give a rodents red butt how long the burner is or was running. You wish like heck that some rocket scientist had not turned the thermostat down that night when the wind was gusting to 70, all night long.

That's usually the same person that will drag race you for position at the light and will burn more gasoline in that one short burst of speed to beat someone to the merge lane than they would ever use during peak demand.

If you want to be efficient, get ODR. If you want to waste money, futz around with the firing rate of youyr heating equipment. It won't be time and money well spent. I've tried about everything one can try. What little I saved wouldn't buy me breakfast.

And if the house is cold, the sound of my wife complaining about being cold and wanting to be in Florida will send me running to the thermostat to set it higher and asking, "Is this high enough?".

I found long ago with my old Bachrach Wet Kit, that there was a place of deminishing return when downsizing boilers. Sooner or later, it went the other way. And if it is 32 degrees and blowing rain and waiting to snow. I'm not obsessing about how long my burner runs. Its nowhere near design. As long as I'm warm, and my wife isn't complaining, life is good.

I once had a Ford Fiesta, 1980  model. It was a 4-banger that would go from zero to 60 in about 15 seconds. Maybe 30 seconds to 80 MPH. But it would never go faster than 80 MPH. And it got 44 MPG. My heating system is like that car. Real quick but only went as fast as I needed it to go.

Robert_25

ODR

Icesailor, your mention of ODR savings vs. reduced firing rate rings true with my observations this winter.  For a few days I doubled the firing rate of my boiler to see how it would behave...it ran almost exactly half as much.  During November and most of December I experimented with running the boiler at 140/160 rather than the usual 160/180...that made more of a difference than scaling back the firing rate, and the reduction in pipe expansion noise was nice as well. 



The only issue I have with ODR is that it can't tell when the wind is blowing.  With a home like mine that sits out in the open, I would have to allow a "wind buffer" in the ODR curve.

NRT_Rob

depends on what heat loss you're calculating

If you are trying to calculate your peak loss... for water temperature calculations, emitter and boiler sizing... you will never get closer than the calculation. the only thing that throws it off significantly is internal heat gain.



If you are trying to calculation yearly energy loss, now that's a lot more complicated.



Thermal mass is only helpful for rapid temperature changes which is not what most heating load calcs are concerned with.

icesailor

Automobile Cruise Control:

Mark E. and I had this discussion a while back. "I" say that there needs to be a "cruise control" for heating systems. When I am driving to Florida in my 2001 BMW 325XI wagon with 130,000 + miles, going down '95 with the cruise control set at 75 MPH, the car drives at 75 MPH no matter what I do unless I step on the brake, gas pedal and click "Off" once. If I am going down an incline, the gas drops back. If I go up an incline, the gas goes up. The brain doesn't know why, but the speed (temperature) is dropping or rising. So, the computer takes appropriate action.

They figured this all out for cars. Why not on heating systems? ODR should and do this as long as there is a thermostat connected and you don't go below the design curve.

Simple logic.

LarryC

Cruise control analogy

Ice,



Your heating system has cruise control !  That is what the thermostat does.  When the house is below the setpoint, it turns on the heat.  When the house is at or above the setpoint, it turns off the heat.  It maintains the house at or around the setpoint.



I think what you want is a heating system that controls the rate of heat being added to the house depending on how far away from the setpoint the current temperature is.  You could have a quick responding system if you had a high firing rate system, but then people would be complaining about short cycling.



ODR curves, and high mass storage are ways to limit the temperature variations.

icesailor

Controling the cruise:

Not so Larry.

ODR is the ultimate cruise control. The theory is that the circulator should never shut off and hold the temperature just at the set point and never shut down. As the temperature goes up with standard ODR, the circulator stops. Ir remains as the temperature and system temperature goes up. But, when the outdoor temperature goes down, the system temperature goes up. It follows the curve, In an automobile in cruise control, it doesn't matter how high the hill is, the engine tried to speed up and the transmission downshifts.

If you have a system like fin tube baseboard, a low mass system, you need to be able to tell the system temperature to rise with a drop in outside conditions. Or set back thermostats coming on. Because that is the same as lowering the outside temperature as much as the set back is.

If you used a 4 way mixer like the Taco "I" series, you can let the boiler loop run at what you want. With a controller, the 4-way controller controls the system temperature.

Gordy

Degree days

Like Boiler Pro. If you have that option the longer the period the more accurate it will be. Obviously for new construction you don't have that option.



 Gordy

Gordy

Degree days

Like Boiler Pro. If you have that option. Obviously for new construction you don't have that option.



 Gordy

Gordy

Ice

 I agree totally. It happens more then a hand full of times. Seen it get 3 times lower then design temp for my area -9 design temp. This historically occurs every decade some time for about a 2- 3 week period in Dec. Jan, or both months together. Could prove to be a bad day in those times when designing on the edge. Thats why heat load programs have that built in fudge factor.



Gordy

SWEI

design days

Been a bit of a moving target here, as there's no official ASHRAE data and the official weather station is our county airport (15 miles south, lower in elevation, and sitting on a huge plain with no impediment to wind from any direction.)  After conferring with two local PEs and a handful of old-timers, we concluded that 13F was "probably about right" and have designed to that OAT.



One morning last January a friend measured -14F outside his kitchen window.  Good thing these old brick buildings have mass!

Gordan

Not exactly analogous to cruise control

Cruise control directly monitors the one thing that it is supposed to maintain: vehicle speed. Outdoor reset monitors outdoor air temperature but it is supposed to maintain indoor air temperature (as a proxy for occupant comfort.) What would be analogous to ODR is a cruise control that measures the pitch of the car (slope that it's going up or down) and controls the throttle on the basis of that. Of course, it would not be able to compensate for different numbers of people or luggage in the car and, like ODR, it wouldn't compensate for (head/back)wind either.



Indoor feedback attacks the problem more directly, and is more analogous to a cruise control. Of course, it has its own issues, because the power to mass ratio for a car is much much higher than for a house heating system so the car responds much much more quickly to inputs, allowing for a simple feedback loop. Cruise control can also rely on engine braking - backing off on the throttle will not just cause the car to coast but actually to be actively slowed down - which you don't have with a boiler. Also, there's only one vehicle speed to monitor... but many indoor temperatures. Picking the right (z)one to be the "master" can be tricky.



Yeah, I know... not saying anything new. :-)

chapchap70

Pull an all nighter on design day.

I would do an actual heat loss on my house as follows:



1)  When weather reports say that the low temperature would seem that it was at or below design day, I would set all zones for 68 degrees during the day.



2)  At around midnight, I would record the temperature in all zones and raise the thermostat in all zones to 75 degrees.  Then I would take a one hour nap near the boiler room.  (Hopefully, I would be awakened by the noise level change if the burner shut off



3)  I would repeat step 2 twice, at one hour intervals.



4)  Depending on the temperatures of the zones, I should be able to tell if the zones and the boiler are properly sized based on the temperatures of the zones.  Obviously, if all the zones had lower temperatures without the burner shutting off on high limit, the boiler would be too small.  If the burner shut off on high limit and the temperatures were lower, there would be too little radiation in the zones.  If the temperatures rose, the boiler would be too large.



5)  To calculate the heat lost, I would shut the boiler off for an hour and record the temperature drop in all zones. 



Doing this late at night would negate heat gain by the sun, stove, etc.

icesailor

Vision:

I guess we don't see the same trees that make up the forest,

The thermostat is the speed control. When it "sees" that the speed is rising because of a  drop in incline (rise in temperature), the throttle closes. When the speed starts to drop, the throttle opens and gives more power.

If you set the ODR reset curve, perfectly, and set the room temperature at 70 degrees, and all is equal, the room will be 70 degrees. You don't need a thermostat. But if you install one, and you turn the thermostat down to 60 degrees, theoretically, you have raised the outside temperature by 10 degrees outside. The thermostat will stop the heat. But when you turn the thermostat back to 70 degrees, you have lowered the outside temperature (theoretically) by 20 degrees. You will need hotter water in the system to recover. There needs to be a simple control that will interpret the differential when the system doesn't come back as quickly as it should and when the room temperature comes up, it backs off.

If you drive your car at 35 MPH down an on-ramp to a highway, what would happen if you couldn't change the throttle setting? or say, you were running at 2000 RPM at 35 MPH but would run at 3000 RPM at 65 MPH. If you set the gas at 3000 RPM as you went on to the highway, for a mile or two before you get to 65 MPH? You might want 4000 RPM until you get to 65 MPH and back off to 3000. And then modulate the throttle. That's what heating systems should do, it's not what they do.

I have a customer with a heat recovery unit with P/S piping and ODR. It works this way as long as the system temperature is high enough.

There has to be a way.