Very nice posts and thank you. Let me try and contribute to this conversation.. Here is the link provided by Alan Cole (kudos to Alan) today at the Placer Count building deparment to get the Outdoor Design Temp...do note that Squaw Valley where i am located is in PLACER COUNTy

 http://www.energy.ca.gov/title24/2008standards/rulemaking/documents/pre-15-day_language/appendices/joint_appendices/JA%202%20-%20Weather%20-%20Rev12-15-day.pdf

lots of discussion about accuracy of the calcs, but from what i see there is a reasonably large error bars on the inputs. Some people may not do the homework to make sure that the inputs are correct. Additionally, i am wondering how accurate (and i mean this as opposed to precise) the input needs to be. It is not clear to this layman how important it is to know the number +/- some error. If anyone knows that it might be helpful for deisgners. For example, if a Manual J is off by 10%, 20%, 50% how much does that truly impact the result. Will this drive my utility bills up by y%, or will this result in over cycling of my HVAC and thereby reduce lifetime by Z%. If we can agree that accuracy is ok within a certain tolerance, then perhaps we can get away from academic discussions and get to my question :-) about the key parameters to size the system that i will need to buy and installl this year.

Robert, many thanks to you for updating your model. I may try to educate myself using your tools before i go spend real money engaging a qualified professionsal. it only makes sense that i get smarter before i talk with someone before spending thousands on equipment. thank you all for engaging in this discussion. DANA1 nice input on microclimates and i tend to agree, but wonder if we are now at a point where we can say design temp is X +/- a tolerance...somewhat akin to triangulating on a soultion.  we may know several data points and from those we must infer the correct value and accept the inherent error value. Badger at some point it seems that science advance faster than the reveiwing authorities and we have to be comfortable and confident that the math gives us a solution that will result in a comforatable environement for the homeonwer (me) as well as providing a compelling rationale to building officicals that we indeed know the variables, the accuracy of our calca and that the result is acceptable.. my 2 cents...
Steve

Steve, The long term energy savings and comfort level that can be achieved are the results of the original heat loss/heat gain calculations.Almost all engineering firms that run Title 24 calcs use a program call EnergyPlus.These programs have gone through all type of case modeling.Extensive testing at labs like Berkeley and others have shown where software errors or human input errors give different results.The equipment selection never comes out exactly with the same number as the load calc.Professional experience and a lot of common sense goes along ways.To give you an example,we recently got called out to assist another hvac contractor on a custom home in the hills of Hollywood.The house was 10000 square feet plus, two stories,lots of south facing glass and had 18 tons of a/c equipment.The complaint the customer had was in late August the upstairs hallway leading to the master bedroom was a sauna.The other hvac contractor was baffled by how the 3 ton unit feeding this area just couldn't keep up with the load.Too make a long story short,whoever did the design calc was off on the massive glass load.Our calcualtion showed that 7.5 tons of cooling was needed.Too compound the situation ,the customer refused to install window blinds or shading do to the panaramic view of the ocean.With all that said, the engineer must be as precise as possible and take in considerations such as the drapes,etc to obtain the most accurate results.

Being off by X percent on a manual J is not readily quantifiable into a percentage efficiency loss. the issues are:

1. does it cause you to go over a break point in equipment sizing? this will effect lifecycle and efficiency to some degree, as well as first cost.

2. does it cause you to use faulty assumptions for energy usage when selecting a class of equipment? this could have a large impact. maybe you choose a high efficiency boiler that just simply doesn't make sense for your particular climate/heatload for example.

3. does it cause you to upsize/upgrade your heat emitters, pipes or pumps unnecessarily? this is a significant cost vector. for instance, maybe it makes the difference between plated or unplated joist radiant systems. that's a big cost differential.


In YOUR case, you want to get "kinda" accurate and given the local variance in temps that means doing a little extra homework, as you have. the difference between a 30 and an 8 outdoor design is huge. the difference between 8 and 14, much less. and this variance is linear, so if you run the number for 14, you can refigure for 8 very easily for "what if" comparisons.

Curiously the 99% design temps for Tahoe, Truckee, and Olympic Valley are all +16F (I would have though they'd vary be a least a degree or three, but I haven't compared datasets.)

The effects of cumulative error from successively padded inputs can add up to quite a bit, but they would have to be pretty egregious to materially effect boiler efficiency if the CA Title 24 +10% limits on oversizing are followed. Granularity of sizing is pretty large on the smaller end of boiler sizes, but undersizing from manual-J by 10% from a Manual-J is preferable to bumping up to the next size, if the next size would put you outside the oversizing restrictions of Title 24. In practice most Manual-J calcs aren't NEARLY aggressive enough on factoring things like plug-loads, drapes, sleeping humans, etc. and natural ventilation rates are usually overestimated too. The inherent overshoot from these errors of omission & commission are typically well over 10%, and undersizing by that amount doesn't lead to discomfort.

Even when oversized by more than 75% most modulating condensing boilers will still hit their efficiency numbers in a decent system design, yet even a perfectly-sized boiler a crummy system design can bring the average as-used efficiency to it's knees (even if it's running pretty much on the numbers at the 99% outside design temp, and meeting the heat load at that temp.)

But what Rob says is true- errors of 50% can be pretty expensive on the upfront radiation cost box on the spreadsheet, so getting numbers that are close to reality are important. Oversizing the radiation by 50% yields a modest increase in as-used efficiency, but not necessarily enough to make up the cost difference within your lifetime when the radiation is radiant floor. (The plated/unplated joist bay example is apt.)

The variability in the inputs is interesting. Even among the experts on the thread, we see a significant variability in the assumption around Outdoor Temp assumptions that will impact the heat loss analysis and will then drive different system sizing (and cost). Thank you all.

That said, my architect has a radiant designer he is proposing for these calcs (after i do them as well :-) . If at all possible, can you please advise on a typical price (or range) that i should expect to pay for the radiant design -- basically a 3 story IDF house with 3300 sq feet conditioned (not living), 58' deep, 36' wide, slab on grade for ground floor, wood i-joist floor system.

Thanks again, Steve

If you think there's a lot of variability in Manual-J input assumptions, try the number of variables in running a full radiant design! :-D

Too many "depends"- 3300' of barn slab run as a single zone vs three 1100' floors run as a single zone or as 3 zones or cut up into 4-6 micro-zones per floor with slab on bottom-floor only makes a real difference in how much the design work is involved, and that's before you get into what it costs for the designer to live & work in urban CA vs. exurban CA vs. MN vs. ME, etc.

It's sort of like asking how much does it cost to design a three story 3300' slab on grade house. The design costs can vary by more than an order of magnitude depending on the particulars.