Marine Zone 4 — Wall insulation — addition
Zone: Marine 4
Location: Seattle, WA
House Built: 1915
Square footage: 900 ft2 single story with 900 ft2 finished basement.
Wall height over basement: 12 ft
Current siding: Cedar, Paint in poor condition.
Insulation: not much, some fiberglass bat stuffed in nooks and crannies as they opened walls in the past. And dirty so there is plenty of air gaps.
GOAL: Comfort
During winter my thermostat reads 70F but still feels cold because the cold walls/ceiling suck the heat out of my body. After reading through GBA articles and forums I thought adding external insulation might be a good idea. Current plan of action:
Existing Walls: Sheet rock > 2×4 framing > some insulation, no vapor barrier > sheathing (plenty of air gaps)
New Addition: Tyvek HomeWrap > Two layers 2″ Roxul ComfortBoard 80 (old house, I fear fire) > 2×4 furring strips > fiber cement siding.
Windows: New fiberglass, inside installation.
Not adding ceiling insulation yet. That will come later after I replace electrical wiring.
Does this plan look reasonable, disastrous, wasteful?
Thank you
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Is there any framing or framing cavity insulation on the addition? Or is the 900' of 2x4 framed leaky space what you're adding the comfort board to?
Adding insulation over an air-leaky wall isn't the best investment, since the air leaks can be anywhere. Whether you add insulating sheathing or not, fixing the air leaks is the first order of business.
The existing wall cavities can probably be dense-packed from the exterior with fiberglass or cellulose, either drilling through both siding & sheathing, or popping off a clapboard or shingle for an easier cosmetic fix (if keeping the siding), which would tighten things up (a lot!). At that point you'll have a ~R10 whole-wall wall, before adding any exterior insulating sheathing. Adding just one layer 2" of rigid rock wool would cut the heat loss from the walls nearly in half.
If the existing windows are single panes + clear glass storms or clear glass (no low-E) double panes, weatherstripping the windows and replacing the storms with tight low-E storms can also make a huge difference in comfort. The low-E storm on the exterior raises the temperature of the interior side glass, reducing the convection drafts and raising the radiant temperature sligthly. That might make the most economic sense if you don't have to pull the windows when adding the exterior insulation.
If you're pulling the windows and replacing them, a U0.25-ish double-pane with low-E coatings surface #2 (the usual low-E glass unit) and surface #4 (the inside facing surface glass) can make a major uptick in comfort. The low-E on #4 lowers the surface temperature of the window a bit, but it reflects body-heat and room heat back toward the source, improving the mean radiant temperature (the primary source of comfort) by quite a bit.
If the foundation isn't already air sealed and insulated, the cold floor and drafts could be a major source of discomfort.
User-6933356,
First of all, can you tell us your name?
Like Dana, I'm confused about this addition. Are you planning to add a new addition to an old house built in 1915? Or are you describing an old 1915 addition that you want to fix?
Assuming you are adding a new addition to this 102-year-old house, what stage are you at? The design stage? Or is the addition already framed?
I agree with Dana that air sealing work always comes before insulation work. An old 102-year-old house like the one you describe is a good candidate for blower-door-directed air sealing. Hire a home performance contractor or a weatherization contractor equipped with a blower door and focus on air sealing.
If you are still at the design stage for a new addition, it's odd that you didn't mention whether you will have wall sheathing, or where your air barrier will be located, or whether you will be installing any insulation between your studs. I recommend that you include sheathing (OSB or plywood) with taped seams -- that will be your exterior air barrier -- and that you select some type of insulation to install between the studs.
The approach you describe -- adding all of the insulation on the exterior of the wall -- is more appropriate for retrofit work on an older house than it is for a new addition, which is one reason why I'm confused about where you intend to install the semi-rigid mineral wool panels. Will they be installed on an older wall or a new wall?
My name is Brad. When I go to "My GBA" there's no option to add a user name or personal information, only option is to add a project. So I'm stuck with the weird User-#### handle.
It's just a retrofit to existing walls. The addition is 1) Tyvek HomeWrap, 2) 4 inches total of continuous Roxul ComfortBoard 80 insulation added to exterior of existing walls, and 3) new windows.
I have 4 original single pane wood framed windows and all the rest have been replaced with single pane aluminum framed windows. Intention is to replace all of them with fiberglass double or triple pane -- undecided because I don't know where the sweet spot is on window value.
Dana, thanks for the surface #4 tip. And thanks for the terminology: I want to improve mean radiant temperature.
I will be removing all my siding so it will be easy to drill holes and add insulation to existing walls. One concern is not knowing what I'm filling over -- e.g. old wiring. If I decided to insulate the wall cavity I'd lean toward removing the sheetrock. That way I could remove all existing bat insulation (most likely incorrectly installed, just like it is incorrectly installed in the attic) and I can inspect the wiring. With an old house like this I don't want to do anything without seeing it first.
At present, the finished basement is the warmest part of the house. It looks like they may have framed a wall inside the cement foundation and insulated it. Primary living space above the basement is the comfort problem.
Martin, regarding a blower door air test, at this moment I can tell you exactly where the air would leak. It would flow out of the old aluminum framed windows. When I remove the siding to install the Tyvek and external insulation I will be able to see my sheathing. I could seal cracks and seams in the sheathing at that point. In addition, the Tyvek will also be properly installed and sealed over the sheathing. I could place a membrane over the sheathing to completely air seal the walls but in an old house like this I'd be nervous. Rainscreen over four inches of Roxul over taped Tyvek over my sheathing seems like the least risky approach to insulating the walls of my home. But maybe I'm wrong and that's why I'm asking.
Thanks for your help,
-Brad
Brad,
I'll address your problems in two separate posts. First, concerning your screen name problem: you should follow the directions posted in this article: How the GBA Site Displays Readers’ Names.
If you are still having trouble after reading the article, let me know and I'll see what I can do to help. My email address is martin [at] greenbuildingadvisor [dot] com.
Brad,
The confusion arises because the word "addition" (in construction terminology) means a new room or rooms added to an older house.
You aren't really building an addition. You are performing retrofit work (or implementing energy upgrades) on existing walls.
OK, we've cleared up that confusion.
Here's my advice:
1. I don't think that adding rigid foam to the exterior of your walls increases the fire risk. But if rigid foam makes you nervous, you can certainly use semi-rigid mineral wool instead if that's what you want to do. Just be aware that installing semi-rigid mineral wool isn't as easy as installing rigid foam, because mineral wool is squishier. That makes fastening the furring strips a little trickier -- it's harder to get the furring strips co-planar with mineral wool than it is with rigid foam.
2. Make sure that you do a good job of installing window flashing (including sill pans) in your window rough openings. Your decision to replace the windows is a good one; it gives you a golden opportunity to get the flashing right. But you don't want to screw up these details -- once the windows are installed, it's fairly difficult to correct flashing errors.
3. Your instincts about the location of air leaks in your home may be accurate -- but it may not. Many homeowners notice air leaks near windows, while ignoring air leaks into a basement or at the top floor (through the ceiling into the attic). Basement leaks and attic leaks aren't noticeable -- but they represent the worst leaks from an energy perspective. I suggest that you read up on the issue. Here are links to two articles to get you started:
Air Sealing an Attic
Air-Sealing a Basement
4. As you contemplate different approaches to fixing your walls, remember that you want to pay attention to airtightness every step of the way. If possible, your walls need an exterior air barrier as well as an interior air barrier.
Hi Martin, there's no screen name in the upper right. Here's a screen snip.
You can get to the My Account page and Update Profile link by clicking the user name in the forum posts. I made those changes but don't see my name yet.
Brad,
My guess is that you have to wait 24 hours for the changes to take effect. Let me know if I'm wrong.
Martin,
1) Good to know. Maybe I'll switch to foam. After watching that building in UK burn up I got leery of foam board http://www.telegraph.co.uk/news/2017/06/22/grenfell-tower-victims-poisoned-cyanide-insulation-released/
2) Correct. I replaced one of the old windows. No window flashing at all on the original. But it was easy to add flashing when installing the new window. Although, no evidence of past water leaks either. I'm sure I'll find leaks as I work my way around the house.
3) I'm sure I have attic and basement leaks too. I'll read that article.
4) Do you agree with this statement by BuildingScience.com? "Rigid materials such as gypsum board, exterior sheathing materials like plywood or OSB, and supported flexible barriers are typically effective air barrier systems if joints and seams are sealed." If so, sealing the sheathing should supply the external air barrier on the walls. Ceiling and attic will need to come later. And it looks like I have different options to install an internal air barrier depending on the next steps I take.
Thanks
- Brad
Ceiling and basement air sealing will need to come later, I mean.
Forgot the link to BuildingScience.com https://buildingscience.com/documents/digests/bsd-104-understanding-air-barriers
OK, leaning back toward mineral wool now. Carpenter ants are a problem in the PNW.
"Ants like wet wood because it is soft and easy to chew. Carpenter ants like insulating foam too."
https://bct.eco.umass.edu/publications/articles/controlling-termites-and-carpenter-ants/
Brad,
Q. "Do you agree with this statement by BuildingScience.com? ‘Rigid materials such as gypsum board, exterior sheathing materials like plywood or OSB, and supported flexible barriers are typically effective air barrier systems if joints and seams are sealed.’"
A. Yes. If your old house has board sheathing, you can't use the sheathing as an air barrier, because there are too many seams to tape. You can either cover the old sheathing with new OSB or plywood and tape the seams of the new sheathing, creating an effective air barrier that way; or you can install a layer of rigid foam on the exterior side of the old sheathing boards and tape the seams of the rigid foam.
On the interior, drywall or plaster can be an effective air barrier, but only if the electrical outlets and other penetrations are carefully sealed.
The stated goal is comfort. With only single pane windows (and no storm windows), the existing comfort shortcomings are likely to be PRIMARILY all about windows!
The U-factor of wood sash single-panes is usually about U-1.0, aluminum sash single panes run about U1.2. That is about 10x the heat loss of a cedar clad 2x4/R11 wall per square foot, about 4x the heat loss of UNINSULATED 2x4 wall, and the windows likely to dominate the current heat load numbers of that addition, while dragging down the mean radiant temperatures.
With windows that lossy the temperature of the inner surface of the glass at 40F, is lower than a clear-glass double-pane at +10F, or a code-minimum low-E double pane at -20F. (See: https://s3.amazonaws.com/finehomebuilding.s3.tauntoncloud.com/app/uploads/2016/04/09102418/h00029_05.gif ) The effect that has on the mean radiant temperature of the room (the most important factor for comfort) can be felt 10 feet away. Having glass that cold also generates convective drafts that can be felt when near the window.
In Seattle's climate triple panes would be overkill (unless going for PassiveHouse levels of energy use). Even double-low-E #2 & #4 surface double panes might be overkill, but they're guaranteed to be higher comfort than a code-min window. A pretty good U0.28-U0.30 argon filled double pane with low-E only on surface #2 would be a huge improvement over what you currently have, if only slightly better than current U0.32 IRC code minimum for zone 4C.
Martin, if I run into board sheathing I'll add a barrier. The first thing I did when I moved into the house was seal the inside wall penetrations. Sealing around the sockets, lights, switches was both relatively inexpensive and noticeably effective.
Dana, thank you. My goal is comfort. I didn't realize the role windows played in that regard. I thought in terms of R value rather than mean radiant temperature. And it's true, they do feel cold from 10 feet away. I have cheap blinds up now that provide very little help, if any at all. I'll pay attention to your window recommendations.
Hi Scott, greetings from North Seattle. You ought to visit my house. I'm cruising towards the finish line on a similar project. In addition to our shared goal of comfort, I have other goals: Net Zero (I'm at 95%), no fossil fuels, improved building safety (seismic retrofits, eliminate combustion appliances), minimal embodied energy, and IAQ.
As others have mentioned, the means for getting a comfortable building is easy: insulate, insulate, insulate, air seal, and insulate. Not necessarily in that order, and sometimes you get two-fers like spray foam and properly flashed windows that do both. But the details is where it gets tricky. I got quotes to have my walls filled with dense pack cellulose and they were VERY reasonable. I'd have done it, except...
I ordered 21 new windows and a 55' truck of reclaimed 3" poly-iso. That's enough to twice go around a 2400 sq/ft house and a 1000 sq/ft garage AND cover a 600 sq/ft roof with 3 layers. I'll see your R-21 minimum walls and raise it to R-36 thank you very much. (One minor downside: with R-36 exterior insulation, an IR camera is no longer useful as a stud finder.)
I began a wall at a time: remove siding, fill all gaps and cracks with Dow Great Stuff Pro Wall & Floor, tape all seams (I tried several and greatly prefer Dow Weathermate), installed my z-flashing (I got bent by B&D sheet metal (dad pun intended)) and insect screen at the bottom edge of the rain screen, installed one layer of poly-iso (Allied stocks 4" screws and foam washers), foam fill & tape all seams, install a 2nd layer of poly-iso with offset joints foamed & taped, and screw through 1" furring with 9" screws (www.bestmaterials.com/detail.aspx?ID=16720) to the joists.
(9" screws - 3/4" furring - 6" polyiso - 1/2 sheathing = 1 3/4 penetration into joist). If you're thinking, "3 in. of penetration will take a beastly drill and hardened bits," you'd be right. I had to upgrade my cordless drill and pre-drill the holes.
On walls with windows and/or doors, I pre-constructed 6" bucks with 8" flanges from 3/4" marine grade plywood sourced off craigslist. Then I cut the insulation around the bucks and continued along. Once a wall was thus insulated and furred, I covered it with Tyvek and then flashed and installed the windows. It wasn't the most efficient way, but my family was living in the house so I considered it important to have the holes in each wall plugged before calling it a day.
I started on the North wall (out of sight, good place to learn) and by the time I arrived at the South wall it was mid-summer. What a difference in comfort R-36 insulation adds. I chose triple pane low-e glass for my South and West facing windows. Once those walls were done the daily heat rise in our upstairs bedrooms dropped 20°. What a HUGE comfort bonus.
I had an uninsulated floor over a crawl space so I bought a spray foam kit and sprayed an inch of closed cell foam to the underside of the subfloor and rim joists. Then I filled the cavities with fiberglass and to break the thermal bridges I screwed a layer of poly-iso to the underside of the joists. Cold floor no more. Unlike the walls, I did not tape the seams. If bulk water ever finds its way into the floor, I want it to leak through.
I'm now trimming out the windows so I still have some bucks exposed and some of the flashing details visible.
I've gleaned many of the techniques from GBA and the Building Science web sites so a big shout out to both sources (thanks Martin!).
Matt: Do you have any pictures to share? (Especially any showing the construction details!)
What's holding back the 5% that's keeping you from making 100% Net Zero? Your wall-R is beyond PassiveHouse-worthy, and you probably optimized the windows for solar gain.
Your biggest energy use now is most likely domestic hot water. If you don't already have one, adding a drainwater heat recovery heat exchanger downstream of the main shower can probably push you over the fully Net Zero line. That could be tough to implement if the crawlspace isn't very deep and the shower is on the first floor rather than on a second story.
Brad's goal was "comfort", not Net Zero energy, and though very comfortable, a Net Zero retrofit can be a pretty expensive undertaking.
Any all-electric house in Seattle City Light's service area (even a really crummy high energy use all electric house) is already pretty much net zero carbon. See:
http://www.seattle.gov/light/FuelMix/
Keeping the energy use ultra-low and sourcing zero-carb electricity directly onto the local distribution grid opens up capacity on the grid, and keeps Seattle City Light's need to import power from BPA bounded. The carbon emissions on SCL's grid sources are solely from those pesky BPA imports, the1% coal and 1% natural gas fractions. If the PV is over-producing when they would otherwise be importing power, it's definitely offsetting those imports. But even with BPA import it's 98% carbon-free power.
Matt: How did you source the reclaimed polyiso? I'm in Shoreline, WA and am in the process of adding exterior rigid foam.
MDUBYA:
https://olympic.craigslist.org/mad/d/polyiso-rigid-foam-insulation/6372302700.html
https://bellingham.craigslist.org/mad/d/polyiso-rigid-foam-insulation/6369413196.html
I'd love to share pics but when I try to post a URL to my project page, I get blocked (for two weeks now). I sourced my reclaimed polyiso from insulationdepot dot com.
My project, with photos scattered throughout, is extensively documented at github.com/msimerson/330-NE-193rd-St. Issue #88 documents the first wall. I hadn't yet purchased the 4" screws and foam washers so I was still working too hard (trying to add 2 layers before any fasteners). I didn't tape the plywood seams on the garage walls. I foamed over the window opening and then cut them out with a hand saw. It wasn't easy to keep the blade straight through 6" of foam so on subsequent walls I switched to pre-building the bucks and strapped them into place before installing the iso. I foamed the iso boards into place on the backs and edges.
What's the last 5% hurdle? Besides the grey winter skies? And Seattle's coldest winter in 32 years? And still insulating in mid-December? And running power tools all year? I estimate it was primarily inadequate ceiling insulation. I dropped one (of 3) ceilings this summer and filled it with closed cell foam (R-6 to R-40). But I still have ~ 1,000 sq/ft of ceilings with minimal (2" of rock wool in part and 3.5" of fiberglass in the other) insulation. The heat pumps are Mitsubishi with HSPF ratings of 13 so they're highly efficient but we still have plenty of heat loss. I may hit Net Zero this year (excluding car). In the future, I might install more solar panels instead of more insulation. That will likely depend on incentives, math, and spare time.
My heat pump water heater accounts for 1.3 kWh/yr, or about 10% of energy use. I have all the plumbing accessible in my basement. The HPWH is already very efficient, setting a very high bar for a water heat recovery system to ever pay for itself.
"a Net Zero retrofit can be a pretty expensive undertaking"
That depends on your time horizon. Paying $N00/mo to gas and electric companies every month is also expensive. I did the math and found that for about 10 years, targeting Net Zero costs about the same as the status quo. For any period longer, Net Zero is cheaper. Stated another way, I've pre-paid for 10 years of energy at today's rates, and every year my system produces after 10 is profit.
But your point is well taken, the up-front costs are stiff: I dropped $15k in insulation, a $33k solar array, $15k in upgraded appliances (two heat pumps (4 heads)), fridge, dishwasher, HPWH, washing machine, bath fans, etc.). But the ROI on most of those upgrades is under 7 years. The 30% federal rebate returned a big chunk already. Seattle City Light also has a plethora of rebates for energy efficiency upgrades. The most expensive component (solar array) has a 6 year payback (details in issue #13).
As Dana points out, SCL has one of the cleanest grids in the country (spoiler: hydro + wind). Rooftop solar arrays are highly complementary to our grid: in the summer when our reservoirs (and hydro production) are lowest, I'm feeding 2/3 of my annual production to the grid. During the winter when it's grey, wet, and windy and our hydro + wind peaks, I'm drawing from the grid. Everybody wins.
But electricity is only 1/3 (~10kWh/year) of the average household energy equation. The average home heats with ~11kWh equiv. of natural gas and drives their cars with 15kWh equiv. of gasoline. I was looking for an escape hatch from gasoline since the Iraq war. When the Leaf became available, and it increased our automotive efficiency from ~22% to ~85%, and dropped our car energy budget from 15 to 4 kWh/yr, it was very easy to justify. Especially since Seattle's grid is nearly carbon free.
That left us with natural gas. I wanted to get rid of natural gas for quite a few reasons: I don't want anyone fracking within 100 miles of my drinking water so why impose that on others?, combustion gases inside the home, gas leaks (I've witnessed 3 buildings leveled by a gas leak), old 80% efficient furnace, huge fireplace/chimney I wanted to remove, and of course carbon pollution. I don't have any baseline data for how much gas this house required because I ripped out the furnace weeks after purchasing it. But I can tell you that our entire household electricity consumption (11kWh) is about the same as the average house consumes in natural gas in this region (12kWh equiv.).
The last reason to go all electric is that I'm a Tesla PowerWall away from being able to operate our entire house off-grid. That could be useful after The Big One shakes us about. I estimate the grid will be down a week and natural gas lines will take weeks to be restored.
I bought a 55' truck load of 3" polyiso, as that was cheaper than buying only what I needed wholesale. I wrapped all my walls with 2 layers, gave away enough to do a friends house, and I still have 100 sheets left over. I'll probably use it to insulate my basement and crawl, just to get rid of it.
Matt writes:
"My heat pump water heater accounts for 1.3 kWh/yr, or about 10% of energy use. I have all the plumbing accessible in my basement. The HPWH is already very efficient, setting a very high bar for a water heat recovery system to ever pay for itself."
Just the controls on the unit probably use on the order 1.3kwh/year for an 8760 hour year. That would be about 6.5-7 watts of standby, not the power used by the compressor or blower. That was probably meant to be 1300 kwh/year(?). Similarly, the 11kwh "...entire household electricity consumption..." was probably intended to be 11,000 kwh/year (?). The average house uses a LOT more than 12kwh equivalent (= ~0.41 therms or ccf) in a season for space heating with gas- 410 therms would be more likely.
In other posts I've noted that with an HPWH the primary way a drainwater heat recovery unit would pay for itself is if it avoids having to upsize to an ~80 gallon unit rather than a ~50 gallon unit, in which case most of the cost is paid up front.
"I bought a 55' truck load of 3" polyiso, as that was cheaper than buying only what I needed wholesale. I wrapped all my walls with 2 layers, gave away enough to do a friends house, and I still have 100 sheets left over. I'll probably use it to insulate my basement and crawl, just to get rid of it."
More than "...just to get rid of it..." bringing the foundation insulation up to current IRC code minimum or better would be in order for this house. R15 continuous insulation is now IRC code minimum for climate zone 4C(!), and any house that has 6" of polyiso for insulating sheathing for above grade framed walls needs at LEAST 3" on the crawlspace & basement walls. If the basement and crawlspace aren't insulated, that would account the lion's share of the remaining heat load, comparable to or even higher than window losses, depending on the the particulars.
"As Dana points out, SCL has one of the cleanest grids in the country (spoiler: hydro + wind). Rooftop solar arrays are highly complementary to our grid: in the summer when our reservoirs (and hydro production) are lowest, I'm feeding 2/3 of my annual production to the grid. During the winter when it's grey, wet, and windy and our hydro + wind peaks, I'm drawing from the grid. "
I don't mean to be harshin' on the mellow here, but....
... hydro production in the PNW typically peaks during the late-spring & early-summer snowmelt runoff, not during the cool and rainy winter season. The watershed areas behind the dams really count on "seasonal storage" in the form of snow, but it's locked up until the temps are warm enough to melt, and it's from mid-summer on into the heavier rains of fall or even early winter where there is sometimes a shortfall in hydro production in that region. Across the BPA sources, first day of summer has typically 2x the hydro production of the last day of summer, and ~1.5x the output of cool-rainy December through February timeframe:
https://www.eia.gov/todayinenergy/images/2012.05.29/DailyHydroBPA.png
Seattle City Light's resources will have a similar profile, so your PV production is really peaking about the same time as hydro production, give or take a month.
Your PV output peak in Seattle may actually be in April or May most years, when days are already long but temps are cooler than in June/July, with less temperature derating of the PV. Cloud cover varies a lot from year to year though- I'd have to look up the 25 year cloud cover & temperatures to better estimate the peak PV season for your area. (In New England it's typically peaking in March, sometimes February, due to an abundance clear-cold winter sky maximizing the PV efficiency.) But it's a fairly flat peak from about half past April through the beginning of August for fixed tilt arrays, according to this resource (the picture on the right):
https://www.homepower.com/sites/default/files/articles/ajax/docs/7_HP131_pg50_Sanchez-4.jpg
But no matter, the summer season is a good time for the PNW to be exporting power to the southwest to help cover the air conditioning load, since the much higher temps are cutting quite a bit into their PV output during the June-July time frame. It's also a GREAT time to be charging up the EV for any summer vacation/weekend driving excursions.
Correction: as Dana points out, my units are wrong. I reduced the kWh values but didn't update the labels. Every kWh label should be MWh.
On DWRs: a 50 gallon HPWH has been quite sufficient for our family of four. I remain skeptical that a DWR is worth my time or money.
"If the basement and crawlspace aren't insulated, that would account the lion's share of the remaining heat load."
The majority of the remaining heat losses are the 600 sqft uninsulated slab and stub walls in the conditioned basement, the uninsulated walls in the semi-conditioned basement, and the under-insulated portions of the ceiling I haven't touched yet. Roughly in that order.
Dana: "hydro production in the PNW typically peaks during the late-spring & early-summer snowmelt runoff, not during the cool and rainy winter season. The watershed areas behind the dams really count on "seasonal storage" in the form of snow, but it's locked up until the temps are warm enough to melt.."
My bad. I used the term "production" when I was thinking of "capacity." BPA cares about peak hydro production because they own transmission lines and can ship the excess. SCL doesn't own transmission lines so we dump water that won't fit in the reservoirs over the spillways. To a point, because if we dump too much it gasifies the rivers and kills the fish. My point was that SCL's summer hydro capacity is limited primarily by our two reservoir systems. When SCL sees peak hydro is weather dependent and more variable than the BPA 5-year moving average suggests.
Six of our hottest years on record are in the past 10. During a low snowpack year (warm, dry, or both) like 2015, peak hydro can shift MONTHS earlier to February and we saw our annual hydro lows in July and August. Even in dry years there are minimum flow rates required for the rivers. Because fish. If producing power in July means not enough water in September, then July production is curtailed. You can see that in SCL's monthly resources chart and water discharges from SCL's two hydro projects:
data.seattle.gov/dataset/Monthly-Seattle-City-Light-Resources-Chart/fjv7-9qhm/data
Skagit River: http://www.seattle.gov/light/Skagit/
Pend Oreille River: http://www.seattle.gov/light/Boundary/
Skagit River Water Discharge by month: waterdata.usgs.gov/wa/nwis/monthly/?referred_module=sw&site_no=12178000&por_12178000_149338=1179899,00060,149338,1908-12,2017-06&format=html_table&date_format=YYYY-MM-DD&rdb_compression=file&submitted_form=parameter_selection_list
Pend Oreille Water Discharge by month: nwis.waterdata.usgs.gov/wa/nwis/monthly/?referred_module=sw&site_no=12396500&por_12396500_149573=1180131,00060,149573,1952-10,2017-09&format=html_table&date_format=YYYY-MM-DD&rdb_compression=file&submitted_form=parameter_selection_list
After the 2015 drought we had record snowpacks that winter and GREAT skiing. But then we had record meltouts starting weeks earlier than normal in April '16. In May the reservoirs are nearly full so even though we had great snowpack, the early meltout caused much of our summer capacity to be lost. Warmer winters (less snowpack) and early meltouts are the trend, and longer summers means more days over which to stretch our fixed hydro capacity. I'd wager if we found a BPA hydro production graph that's six years newer there'd be a pronounced leftward shift in peak production.
During those longer warmer summers, when our reservoirs are essentially a discharging battery, I'm feeding the grid and reducing summer demand. In the winter months (Nov-Mar) when I'm drawing from the grid SCL has reliably sufficient hydro and wind capacity. Enough that by December we're typically dumping water over spillways and idling wind turbines. See water discharge above. That's when I'm drawing from the grid.
SCL has been a strong proponent of solar for many years. They actively support community solar projects, have booths every year at Northwest SolarFest (I got the "solar is complementary to SCLs grid" idea from an SCL employee), and contrary to my expectations, they were extremely pleasant to work with while getting my array deployed.
Solar is also complementary to SCL's grid in another way. We're (SCL is city owned) buying solar power from California (during the day, when prices are free or negative) and then selling our hydro power later in the day when prices are [much] higher. sccinsight.com/2017/05/21/seattle-city-light-join-energy-imbalance-market/
"Seattle City Light's resources will have a similar profile, so your PV production is really peaking about the same time as hydro production, give or take a month. Your PV output peak in Seattle may actually be in April or May most years"
In May and June most mornings are overcast with chance of rain, burning off by mid-day. Then summer arrives on July 8th. The clouds depart and the rains finally stop. My monthly production for 2017 (kWh): 228, 337, 547, 929, 1250, 1370, 1640, 1310, 904, 493, 163. I peaked in July and my actual production nicely correlates with PVWatts predictions for an East-West facing array with an 18° pitch.