There's some really good information on window maintenance and repair here as well (obviously haven't read all of it, or even most but looks pretty interesting):
http://historichomeworks.com/forum/viewforum.php?f=16
The Andersons we had were the "swing to the outside" casement version. It was a pretty hot area in the summer and we couldn't get them open far enough to really allow the breeze to blow straight through (they'd go to maybe 60d instead of 90d or more). The swing ins look like they'd alleviate a lot of those problems, so definitely filing that away for future thought..
You're certainly correct on the ease of cleaning aspect of the tilt in windows., especially on the upper stories - worth keeping in mind!
I hear you on a lot of the older house issues - when the wind goes in one side and out the other it doesn't matter how cool the hand axed logs look
Having said that there is imho a lot of trade offs between short term and long term maintainability that make the whole conversation .. complicated. Its certainly true that some older stuff was well made and much was most certainly not and a lot of the stuff that was could certainly use a lot of repair and maintenance at this point.
On replacement costs I'm still having some difficulty figuring the savings for the general case. For new windows it seems to generally make sense to get higher E windows (although the math on what specifically is better is complicated there as well, and depends on the window types/environment). If you have giant picture windows in a cold climate then likely new window there is a no brainer from a comfort perspective but just figuring for "normal windows" I'm not convinced
I'm looking at the U values (measure of heat transfer rate) for various glazing types (for example:
https://www.wbdg.org/resources/windows.php about half way down the page has a nice chart that seems to line up pretty well with other data I'm finding such as
http://www.aspenpitkin.com/Portals/...ode/1993 ASHRAE Fundamentals Ufactor only.pdf) and the best glazing types seem to be around 3x better than single pane (until you get to the crazy stuff, but that's only the center of the pane U factor the overall window construction matters a lot as well).
Using exceedingly rough math we can get some completely bogus cost saving estimates (hows that for confidence
). Most window mfgs publish some sort of U values so it would be better off plugging those in and doing the math for any specific house, the example I'm pulling out here is only very roughly illustrative. Baseline numbers I'm using are (rounding slightly) single pane U of 1.06, double pane U of 0.5, triple U of 0.3, and best in class might have a U of 0.1 or so. I'm figuring window heat loss should be 20% of the total house heat loss as an extremely high guestimate assuming that the house is somewhat normal construction (20% might be a baseline for a house with a LOT of single pane windows in really poor maintenance, normal is likely closer to 10%). In theory if storm windows were 100% effective you ought to be really close to or perhaps better than double pane, of course the enclosure likely leaks more than a well installed double so say U of 0.75 or so (imho that's highly pessimistic but I'm biasing the data against my assumptions, the ASHRAE numbers I found penciled out closer to ~0.55U or even less depending on the storm window type).
Using those numbers to start with we'll assume that ~20% of the heating/cooling cost is going out the window with single pane and no storm windows as the baseline with a somewhat arbitrarily chosen yearly heating/cooling bill of $1000 to make the math easy.
Window | Estimated U | savings over baseline | yearly energy savings<br>(arbitrarily chosen $1000 baseline) |
Single pane no storm | 1.06U | 0% | $0 |
Single pane with storm | 0.75U | 5% | $50 |
Double pane | 0.5U | 10% | $100 |
Triple pane | 0.5U | 14% | $140 |
Crazy reflective low E super triple pane with some sort of exotic gas | 0.1U | 18% | $180 |
Assuming ~12 windows (which is what we have in our house, not counting the sliding glass door which is an energy pig as well) and a cost of $500-1000 per window (depending on how fancy you get, including installation for that price..) that's $6,000 to $12,000 to replace all of the windows.
So even at the best case where I managed to use get the 0.1U windows for $500 per (unlikely) its still 33 years to pay that back
If we assume a more average cost of say $10k for the triple pane at $140/year that 71 years to payback. Even if you could get the 0.1U windows at $250 (extremely unlikely) its still over 15 years...
So unless your heat loss is way higher than I'm figuring (certainly possible) or there are other confounding factors (also possible), the math on replacement versus repair gets kind of hard to figure imho; especially given the expected lifespan of most windows (less than the payback time looks common - I'm seeing failure estimates for the seals in double/triple pane windows in the 25-35 year range).
A somewhat more pessimistic study can be found here:
http://www.oldhouseguy.com/assets/StudyReplacementVsHistoricWindow.pdf - imho its a bit biased against the replacement numbers but still its interesting to walk through.
This is all at odds with what the window salespeople claim:
http://www.efficientwindows.org/energycost.php - I can't figure their math though.. the only way it seems to work is if they're figuring the % savings on the windows against the whole house cost (but maybe I'm just doing it wrong - in which case someone please say so
). Supporting my math:
http://www.traditional-building.com/Previous-Product-Reports/3-windowsJune2008.htm - more support:
https://heritage.utah.gov/history/windows (even figuring 25% total savings the math wasn't very good there).
I'm also discounting some classes of comfort issues which can be worth the expense
If the window feels cold (especially to the other interested parties) the math starts to matter less