Musings on the "Design Day"

Jamie Hall · January 4

The effective heat loss rate per degree F is one of the intermediate steps of Manual J, so you can extrapolate from one temperature to another, assuming it's linear — which it really isn't, but… close enough.

Various folks have mentioned wind, and wind effects are one of the really big variables which virtually no heat loss calculation properly integrates. In must situations, the various allowances and fudge factors involved take care of it, but in exposed locations (hill tops, coast lines, anywhere on the Great Plains outside of cities) a good stiff breeze can close to double the building heat loss — even if it is all nicely tightened up and not draughty.

Intplm. · January 4

I am finding this conversation very interesting.

In reading the post it always brings back a thought that directly effects our industry, and that is the design temp. of dwellings/buildings. That temp being 72deg.

I have often thought of the painstaking trial and error that brings us to a practical working conclusion.

@Jamie Hall Nice that you have time on your hands. 🙂

ChrisJ · January 4

https://forum.heatinghelp.com/discussion/comment/1831877#Comment_1831877

I don't think it takes the chimney effect into consideration on multifloor buildings either.

Kaos · January 4

Wind can definitely hurt. Lets put some rough numbers to it. Calculating wind effect is not easy so all this is very rough ballpark.

Typical wind load is around 5Pa, lets say exposed it would be 10Pa. Assuming a 2000sqft two story house, 8' walls. Assume 0F outdoor temp.

Typical code min (that is 3ACH @ 50Pa), that house without wind would leak 2000*8*3/60=800CFM at 50Pa, under typical stack pressure (Nfactor of 20) that would be 800/20=40CFM.

40cfm*1.08*70Fdelta=3000BTU, so near squat.

Under heavy 10Pa wind, that jumps up to around 160CFM.

160cfm*1.08*70F=12000BTU, defiantly something but not a "lot"

Leaky balloon frame house would be around 20ACH@50Pa. That works out to 266CFM stack pressure leakage. Heavy wind 1000cfm.

No wind:266CFM ~20kBTU

Heavy wind:1000CFM ~75kBTU

Both of those numbers are definitely real heat loss. It also means wind is practically howling through the house, there are much easier ways to fix that than by throwing more heat at it.

Also no amount of heat will make a house that leaky comfortable.

fentonc · January 4

Wind, sunlight, thermal mass, etc. all add variability to the heat loss of a house, so I was interested in trying to quantify how much the temperature-dependent heat loss (so BTUS/HDD65-hour) of my house varied over time. I made a chart showing BTUS-delivered/HDD65-hour vs hour of the day for a week or so, which I thought was pretty interesting.

The house has a lot of southwest facing windows, so you can see the strong effect of incoming sunlight (sometimes dropping the delivered BTUs to zero). Looking at midnight (the left edge of the graph), when the sun definitely isn't shining, we see the heat loss still varies from 300 BTU/HDD65-hour to 450 or so. The actual temperature during this window varied between 14F and 40F, I think, with most of it being in the high 20s, low 30s.

The 1st floor thermostat was set at a steady 68F, the basement was set at a steady 62F, and a thermal purge scheme dumped residual heat after all of the thermostats had been satisfied into the basement, 1st and 2nd floors (the 2nd floor still has the thermostat off, so only ever receives residual heat or heat via conduction/convection from the 1st floor).

DCContrarian · January 4

https://forum.heatinghelp.com/discussion/comment/1831907#Comment_1831907

That analysis looks spot-on and I agree completely with the premise that the better sealed a house is, the less of a factor the wind is. Some anec-data: Right now in DC the NWS is saying it's 29F with 17mph sustained winds gusting to 40mph. It's pretty blustery out. My house is new and relatively tight, it blower door tested at 1.4 ACH50 (although people serious into high performance construction would scoff at that.) My design temp is 22F (99th percentile), so I'm 7F above design temp.

I have three thermostats in the house, and right now, 12:20 PM, all three are off. The sunshine coming through the windows is enough to warm the house. All three are set for 70F, they are reading 70.7, 70.0 and 71.6.

If you've never lived in a tight house, it's a very different experience.

DCContrarian · January 4

https://forum.heatinghelp.com/discussion/comment/1831913#Comment_1831913

This is really good analysis. The only question I have is how you're measuring delivered heat.

I'm a big fan of using actual fuel usage to estimate heating load. The assumption that underlies Manual J is that BTU per heating degree-hour is constant. Clearly that assumption isn't true, but the economists like to say "all models are wrong, some are useful," and it has proven to be a useful assumption.

This graph shows one of the pitfalls of using fuel usage, which that if a house has significant solar gain just taking seasonal fuel usage and dividing by seasonal degree-hours (degree-days times 24) will lead to an estimate of heat loss that is too low.

I'm curious about the periods of zero delivered heat in the evening, in the 9, 10 and 11 o'clock hours. Any thought about what happens then? It's what I would expect with a setback thermostat but you said you had constant setting.

fentonc · January 4

@DCContrarian - I have 3 zones of fin-tube baseboard. I measured the linear footage of fin-tube for each zone, and I measure the supply and return temperatures for each zone to get the average temp. I used a table provided by the baseboard manufacturer that shows heat output for various water temperatures with 65F input air. I do linear interpolation using the table, the average water temp and the current air temperature in the room, to get the estimated BTU output at any given time. About once every second, I accumulate the BTU output for any zone where the zone valve is active.

It's not perfect in terms of accuracy, but it's pretty good in terms of precision / repeatability. The actual data is reported for BTUs delivered in the last 60 minutes at every hour:30, and then averaged using a moving 3 hour window.

fentonc · January 4

@DCContrarian - re: evening hours with zero heat - That was actually a warm front where the temp got up into the 50s, so you're mostly seeing the boiler 'pulsing' heat into the system since it's such an oversized CI boiler. Here is about 3 days worth of hourly BTU/HDD-hr + actual outdoor temperature data around that event (starting at noon):

DCContrarian · January 4

What's interesting is that looking at that graph, on an hour-by-hour basis it seems that there is zero correlation between temperature and heating load.

jumper · January 4

If you want to think about something while going to sleep; then there are other considerations to sizing heating system. One is how fast you want your building to warm up to what. Steam radiators beats HHW baseboards in that department.

Jamie Hall · January 4

I think what you may be seeing in that graph is an illustration of the "flywheel" effect of heavy cast iron radiation.

fentonc · January 5

@Jamie Hall - it's all the 'regular output' aluminum fin + copper tube baseboard.

Jamie Hall · January 5

https://forum.heatinghelp.com/discussion/comment/1832032#Comment_1832032

well it was an idea

fentonc · January 5

@DCContrarian - If you just look at actual BTU/hr vs outdoor temperature, the correlation is a lot better:

DCContrarian · January 5

My guess is if you just took the nighttime hours the fit would be even better. Remove solar as a factor.

fentonc · January 5

Looking at midnight to 7am, it is better:

GGross · January 6

https://forum.heatinghelp.com/discussion/comment/1831852#Comment_1831852

Using a software calculation you just go to the design parameters and change the design day temperature. Most come pre-loaded with data from ashrae, and also let you manually enter values, so you just change the "0" to "-10". approximately 15 seconds to accomplish if you are familiar with your software

Musings on the "Design Day"

Comments

Categories