Energy Efficiency and Economics at Walnut Mews

Our condo HOA had a meeting last fall, and somebody brought up selling the flat plate collectors on the roof that are part of our defunct solar thermal hot water system.  The 750 gallon cylindrical storage tank rusted out in 2003 after 20 years of service.  The outbuilding that houses it was basically built over the tank, so swapping it out for a new one would have meant either chopping the thing up in place with a cutting torch and building a new one on site, or removing the roof, which nobody was keen on.  Some plumbing got re-routed and the tank sits there still, derelict.  It was also mentioned that the main boiler for our hydronic district heating might be nearing the end of its days.  I volunteered to look into whether it would make economic sense to repair the solar thermal system, and what the options were for the boiler.

Given that flat plate solar thermal collectors generate an average of about 1kBTU worth of heat per day per square foot (according to the US EIA), and given that we have about 250 square feet of collecting area (nine 28 square foot panels), the current system ought to collect something like 250kBTU/day.  Our current boiler consumes 520kBTU/hr worth of gas, meaning that the solar thermal system could at best displace a half hour’s worth of operation each day.  Gas costs about $8/million BTUs, so the boiler costs about $4/hr to run.  If we assume optimistically that system losses are negligible, and that the boiler runs at least half an hour a day 250 days a year (it was only hooked up to the baseboard heating, not the domestic hot water) then the solar thermal system is capable of displacing something like $500 worth of gas each year.  This is a best case scenario though, since the hydronic system needs water that’s hotter than the flat plate collectors can make it (so the boiler will have to do some work to boost the temperature) and because the system losses are almost certainly non-negligible.

Still, $500/year might be a significant savings.  To know whether it’s really worthwhile, we need to know how much it will cost up front to get this savings, and how long we ought to expect to be able to collect it (i.e. what’s the system’s expected lifetime).  I got wildly varying estimates of the cost to get the system up and running again.  At the low end it was $5000, to leave the rusty tank where it is and put a collapsible storage bladder in the crawlspace.  At the high end it was $20,000 to remove the old tank and build a new spray-foam insulated stainless steel one in its place.  I used this calculator to sanity check my energy numbers above (which don’t seem crazy), as well as the estimates.  It suggests that all in, the total system cost including installation would be something like $28,000.  I suspect that a plastic bladder in the crawlspace wouldn’t be as efficient or as durable as the new stainless tank.  For the sake of argument, let’s say the cheap option will only last 5 years, and the expensive one will last 30 years.  The original tank lasted about 20 years.  Here’s what it looks like today:

Derelict Solar Thermal Storage Tank

PACE and Payback Times

So which of these options makes the most sense?  How do they compare to replacing the existing boiler with a high efficiency one, or doing nothing at all?  What if we have to borrow money to make it happen?  What if we have to decide between doing the upgrades and putting money into our retirement savings?  For some reason, the most common financial metric that seems to come up in evaluating energy efficiency is the simple payback period, which is just the capital cost divided by the resulting savings per unit time.  So for example the $5000 option, saving $500/year, would have a payback period of 10 years.  But simple payback is actually a pretty poor metric, for at least two reasons.

First, it ignores the time value of money.  Would you rather have $100 today, or $100 ten years from now?  Simple payback implies that you’d be indifferent, but pretty much everyone would rather have the money now.  Second, it’s not easily comparable to other investment options.  I suspect it’s popular because it’s easy to understand  (most homeowners can hopefully do division) and also because people are worried about abandoning their investment before it’s paid off.  If the payback period is 10 years, and you suspect you’ll be moving in the next five years, then you won’t get your money back.  In an ideal world, whoever was buying your house would figure its energy efficiency into its value, and you’d get paid that way, but unfortunately this isn’t the way the real estate market works.

Property assessed clean energy (PACE) financing programs attempt to address this market failure.  A city or other jurisdiction issues bonds to create an initial funding pool.  That money is then loaned to homeowners who want to make energy efficiency upgrades to their dwellings, but instead of the loan being tied to the person, it’s tied to the property, and repaid with a property tax assessment, so future owners who benefit from the efficiency measures pay for their share.  The programs are usually designed to be revenue neutral, with the full cost of servicing the bonds being covered by the additional property taxes collected.  Municipal bonds are generally tax-exempt, property taxes are backed by the value of the property itself, and the government’s well-established ability to enforce collection, so the financing ought to be cheap.  Unfortunately these same qualities have also led the Federal Housing Finance Agency (which runs Fannie Mae and Freddie Mac) to consider PACE assessments liens against the properties they’re tied to, in part because the energy improvements supposedly provide “no community value” unlike property taxes that support the upkeep of sewage lines, schools, etc., and in part because they allegedly increase the risk of default to mortgage lenders.  Since most mortgage lenders won’t touch loans that aren’t compatible with Fannie and Freddie, this policy has effectively killed PACE financing.  The NRDC and others are suing to get the decision reversed.  A legislative solution died in committee last fall… and I can’t imagine that the current congress will do any better.  I’ll return to PACE financing below, but for now let’s assume that leaving your investment behind isn’t a problem since in an efficient market, that would be the case.

NPV, IRR and Discounted Cashflows

Usually business investment decisions don’t use the simple payback periods.  If you want to compare the relative attractiveness of several potential investments, the internal rate of return (IRR) or modified internal rate of return (MIRR) make more sense, because they take into account the fact that, for instance, you’d rather have $100 today than $100 ten years from now.  In real life available investments may be mutually exclusive, and the scale of those investments with highest rates of return may not be large enough to absorb all the capital you’ve got on hand.  If you have $10,000 to spend on energy efficiency, and you have to choose between investing $1,000 at a 20% rate of return, or investing all $10,000 of it at a 10%, clearly the latter option is better, even though the rate of return is lower, because the total value ends up being greater.  For this kind of calculation, we want to use something like net present value (NPV), which adds up all the future cashflows resulting from a project, and discounts them to the present.  This discounting is what takes into account the fact that we generally prefer our money now, not later.  If you’ve got $100 on hand now, and you can invest it with no risk at 5% for a year, then $105 next year is worth $100 today.  The same holds true if someone is willing to lend you $100 today, in exchange for your promise to pay them back $105 next year (a 5% interest rate or cost of capital).  In the context of our putative cheap option solar thermal project, the NPV and discounted cashflows involved look like:

NPV = -$5,000 + $500*((1+r)-1+(1+r)-2+(1+r)-3+(1+r)-4+(1+r)-5)

where r is the applicable discount rate.  If we’ve got the $5,000 on hand, and it would otherwise be going into a money market fund earning 3%, then r=0.03, which gives us an NPV of -$2631.  This is not a good investment.  It’s even worse if we have to borrow since then the discount rate is probably going to be higher.  If we borrow $5000 at 7%, and pay the loan off in 5 years, we have to pay back $1219/year to get a savings of only $500/year, and five years of -$719/year cashflow has a NPV of -$2950.  We can also ask how many years the system would have to last for to be worthwhile.  At what point does the annual savings start to make up for the investment up front?  How many years do we have to consider before we have an NPV > $0?  In the case of borrowing the $5000 up front at 7%, if we keep the loan term and lifetime of the system equal, the investment doesn’t make any sense until it lasts for 18 years.  This is analogous to the simple payback calculation… but 18 years is a lot longer than 10!

If we consider the more durable and expensive stainless steel tank option, and instead of hooking it up to the baseboard heating, which only gets used for part of the year, we use it as a pre-heater for the domestic hot water (DHW), which gets used all year long, we’ll save more like $730/year on gas, with $20k in setup costs. If this is financed at 7% for 20 years (which is about what Boulder’s ClimateSmart PACE program would have given us), on a system that will last for 30 years, we end up with a NPV of -$10941.  It looks like we could potentially knock $3000 off that with a rebate from the county, but that’s still significantly worse than the cheap option.  There’s also up to $1500 in tax credits available from the Feds, but since our HOA is a non-profit organization, it wouldn’t apply, and even if it did, all of these possible “investments” are still money losing propositions.  So at least with the cost estimates we’ve gotten so far, it doesn’t sound like renovating the solar thermal system is going to fly.  If it can really save us $730/year in gas, then to make it work under the ClimateSmart loan program (say 7% for 15 years) we’d need a system that would last at least 15 years, with a net installation cost (after rebates and incentives) of less than $6,500.

What about efficiency?

So what about improving the efficiency of the boilers?  Our current 82% boiler (an AO Smith HW520) looks like it costs about $5,500.  A 95% efficient replacement with a similar output rating, like this one from Laars is more like $10,200.  Our DHW currently comes from 2 AO Smith BTR-199 water heaters, each of which looks like it would cost about $5,100 to replace directly.  With DHW the important variable is less the heat output, and more the amount of hot water that can be supplied during peak demand.  Currently we’ve got 264 gallons/hr at 140°F rise.  The AO Smith BTH-300A is 96% efficient and will supply 250 gallons/hr at 140°F rise, and it costs $9300.

Here’s the boiler:

Hydronic Baseboard Heating Boiler

and the hot water heaters:

Domestic Hot Water Heaters

So, unless I’m seriously misunderstanding something, or getting bad prices, there’s no reason whatsoever not to get the high efficiency hot water heater.  For the same hot water output it costs less than the low efficiency models, and that’s without any kind of incentives or rebates.  No calculation required.  It’s worth noting that if we did find someone who could get the solar thermal system up and running for $6500 with a 15 year lifetime, that would let us save some money on the hot water heater, since it wouldn’t have to be as big.  So we might consider working that capital investment savings into the calculation for the solar thermal system.  If it lets us downsize our gas fired system, then it might be worth investing in even if it has a modestly negative NPV.

The situation with the boiler is less clear.  We currently spend about $3550/year on gas for space heating.  Upgrading from 82% to 95% efficiency would save us about $460/year, but getting that additional efficiency would add $4700 to the up front cost of replacing the boiler.  If we expect the boiler to last for 15 years, that cashflow series (-$4700 the first year, and $460 for the subsequent 15 years) has an IRR of 5.2%, which isn’t bad, but if we have to finance the extra $4700 at 7%, then we’ll end up with a negative NPV, since the financing rate is higher than the internal rate of return (NPV=-$510).  On the other hand, if we could somehow manage to get the $1500 federal tax credit, the IRR goes up to 11.6%, bumping the NPV up to $990.  The NPV remains positive even if the hot water heater only lasts 10 years.

But it’s not about the money?

By now the eyes of most environmentalists have glazed over.  It’s not about the money.  Saving energy and avoiding greenhouse gas emissions are their own rewards.  I agree, but even if your primary goal isn’t financial savings, you are still probably financially constrained.  You only have a finite amount of money you can put toward saving energy or avoiding carbon emissions, and so it still makes sense to prioritize your investments according to how effective they are in the service of those goals.  If you want to work emissions into the above calculations, you can pick a price for CO2 and see how it affects the financial outcomes, based on the carbon content of various fuels and the power mix of your local electrical grid.  The solar thermal system above would displace around 5 tons of CO2 emissions per year, so if you priced CO2 at $50/ton the annual savings would go from $730 to $980, which improves the economics (though not enough to make the quotes we got worthwhile).  If you’re interested in avoiding emissions, reducing electricity use is probably the best option, since coal is extremely carbon intensive, our grid has a lot of coal in it, and coal fired power plants are only about 35% efficient at turning heat into electricity, and lose another 7-10% in transmission to your house.  Our condo uses about 3600 kWh of electricity a year, and at a plant-to-plug efficiency of 25% (assuming an all coal grid) that means each year we’re responsible for putting out about 5 tons of CO2 due to our electricity consumption.

There’s also an argument to be made for using less energy to do the same jobs, even if the cost of the efficiency upgrades ends up eating all of the savings, because you reduce your exposure to future price fluctuations.  It’s a hedging strategy.  You can lock in a particular price, and plan around it, instead of being at the mercy of volatile markets.  Natural gas prices in particular have become wildly variable over the last decade.  Volatility is not the same thing as steadily increasing prices though.  It’s popular these days to predict ever increasing energy costs.  Former oil executives, Wall St. analysts, dark green collapsitarians and green building gurus all do it.  It has become a pervasive truism that energy prices are only going to go up, just like home prices in 2006, except instead of the giddy cackling of house-flippers, these predictions are ominous.  I don’t think it’s a coincidence that our apparently financially questionable solar thermal system was built in the early 1980s, when a similar mentality had taken hold.  Like that creepy Boards of Canada song “Energy Warning”, which is a real sample from a 1979 PSA of this vintage:

Or how about this Saturday morning cartoon interlude with Captain America from 1985, which I actually remember:

Then you look at what actually happened (all stats from the US EIA Annual Energy Review 2009):

Coal has no obvious trend:

US Real Coal Prices 1949-2009

And since coal is mainly where our electricity comes from:

US Real Electricity Prices 1960-2009

Oil has certainly been volatile since the Texas Railroad Commission cartel collapsed, but are we really in the midst of an endless increase in prices, or are we just seeing another unpleasant spike, like the one that freaked us out in the early 1980s?

US Real Crude Oil Prices 1949-2009

Gasoline is also volatile, but the real 2009 average price was only 15% higher than in 1949.

US Real Gasoline Prices 1949-2009

Natural gas is the only fuel with a credible upward trend, as it’s made its way into the electrical generation market:

US Real Natural Gas Prices 1949-2009

So, I’m not convinced that we’ll only see rising prices.  Volatile prices, sure.  But you can see how the spike around 1980 would have scared the bujeezeus out of people, given how relatively calm things had been up until the mid 1970s.  I’m not suggesting I think prices are going to go down, or even stay the same.  I’m saying I have no idea what energy prices are going to do.  The US EIA is a great source of historical data, but it has just as lousy a track record as the Wall St. analysts and your average chimpanzee at predicting future energy prices.  I have no reason to think that I can do any better.

Selling efficiency purely on the basis of cost savings necessarily means making a prediction about future energy prices.  Selling efficiency as financial insulation from energy prices seems much more honest to me.  I’ve never depended on a car on a regular basis, so I don’t really know what it’s like to care about gas prices, but it does seem to stress a lot of people out.  Not having to think about it is one way that biking improves my quality of life, and that’s worth something.  However, our capital markets aren’t always so good at encapsulating that kind of benefit.

Other People’s Money

In the case of the PACE financing mechanism, we have to come to some kind of consensus on what qualifies as a good enough investment, because the person making the decision about the investment may not end up being the one who pays it off.  As far as I can tell from the list of eligible measures posted on the website and the recommendations that staff sent to the County Board of Supervisors, this sense of “good enough” is not rigorously codified anywhere.  There’s an enormous list of eligible improvements, including both high efficiency boilers and solar thermal hot water heaters, which as we saw above can end up being very different value propositions.  There’s also photovoltaics, and reflective roofs, high performance windows, attic insulation, heat recovery ventilators, seasonally shading tree plantings, light tubes, and much more.  There doesn’t appear to be a requirement that applicants get an energy audit (though it is recommended) or that the measures undertaken necessarily need to either optimize for financial return or greenhouse gas emissions avoided.  Or that they even need to be net positive financially, or reduce greenhouse gas emissions at all.  So maybe it’s not surprising that the FHFA decided to slam on the brakes.  Or maybe I’m just missing the relevant program details.  I’d like to think that I don’t understand something, because it doesn’t seem to make sense.

If anyone else out there understands the situation better, and feels like explaining it, I’d love to know more.  In the meantime, here’s one last energy related photo:

Outbuilding Roof Insulation and Snow

The far end of the outbuilding houses the boilers for our hydronic heating system and domestic hot water.  You can see that the snow has melted significantly.  It’s about 25°C in there, and there’s no flat ceiling creating an attic space (though I suspect the rafters are insulated, since the snow is doing okay away from the ridgeline).  The near side of the outbuilding is insulated, and contains a “common area” that nobody uses.  It’s probably about 10°C inside.  The building in the sun in the background clearly has poor attic insulation.  The snow has already mostly melted off even though it’s well below freezing outside, resulting in some scary looking icicles!  The snow on our sunny south-facing balcony railing (which you can’t see here) is fluffy and intact.

8 thoughts on “Energy Efficiency and Economics at Walnut Mews”

  1. too bad the tank situation is so, um, idiosyncratic. It seems that a new tank should be significantly less than $20,000, if it wasn’t housed so un-ideally. Of course, free volunteer labor could lower some of this cost, if you could find a group of semi-skilled roofers/carpenters/plumbers willing to go to town.

    It does seem as though you guys should concentrate on electrical efficiency though…

  2. I looked around a bit and found this 325 gallon tank that goes for $4350 is about the biggest one that would fit through the door. And there’s a barely used 200 gallon one for sale on Craigslist in Denver for $1500. Really too small though. Would be a good deal for a smaller residential system though. I talked to some local welders with an arc cutter, and they thought they could get the old tank out in half a day at $75/hr = $300. I wonder if we could get the re-plumbing and insulation done for another $2000, or how long I’d have to wait to find a suitable tank for sale locally at a big discount.

    Ugh, and I know. The BTUs are awful. That’s the one aspect of working on buildings in the US that I’m not keen on!

  3. The Governor’s Energy Office also has up to a $3000 rebate for solar thermal hot water, which if it applies to “orphan” systems like previous rebates did, might mean we wouldn’t need to get a loan at all with the above prices. Say $4350+$300 for the new smaller tank and the removal of the old tank, and another $1350 to get everything plumbed and insulated, would be $6000, minus the $3000 rebate, for a $3000 up front cost. If it really saved us $730/year and lasted 15 years, we’d get an IRR of 23%, and the project would have a NPV of $3400, which would be awesome.

  4. It would be a lot nicer to read your post if you converted everything to metric units at ingress instead of doing math with BTUs.

    How big is the rusted-out water tank you have now? Like, how many food-grade HDPE 55-gallon (200ℓ) drums would it take to replace it? Six? They seem to cost US$25-US$100 new, which is about 12-50¢/ℓ, while that 125psi (860kPa) glass-lined stainless-steel monstrosity you found is about US$3.50/ℓ. 860 kPa is 88 meters of water column, and you probably don’t need more than a tenth of that. I think the 125psi is for containing the output of a hot water heater that might inadvertently boil the water and generate steam pressure.

    If you bought your HDPE 200ℓ drums used, you might be able to pay even less for them, like US$10 or something. Someone reported getting them for US$8 used from their local Coca-Cola bottler.

    Around here, because the water mains are low pressure, everyone has a plastic water tank on their roof, typically about 500ℓ worth. I don’t know how much they cost exactly but I’m pretty sure they don’t cost US$1800 (proportional to that hot-water tank you linked). I’m thinking more like US$60.

    I imagine that if you had a couple of feet between the water tank and the walls of the outbuilding, you could dump in any old crappy insulating material and get near-perfect efficiency. Straw, say, or sawdust, or shredded paper, or discarded clothes. No need to go with fancy, expensive materials like styrofoam when you don’t have to fit inside the thickness of a normal residential wall, right? Just be careful about sparks. (You can get flame-retardant shredded paper, which might be wise and/or legally required.)

  5. Checking out livestock water tank pricing suggests that costs US$270.60, or 24¢/ℓ. That’s 14× better than the hot-water tank you found, although that might be a little too inexpensive — it has a big open top, so it would lose heat.

    As far as insulating with straw, you could probably lay down a layer of straw bales on the floor, set the barrels or tank on top of them, and brick them in with more straw bales.

    I’m not a plumber, so dunno, maybe I’m making some kind of dumb mistake here, but it seems like the figures you had were just crazily high. Seems to me it’d be better to spend extra cash on slightly more complicated plumbing: a thermostatic system to keep the stored warm water at a consistent temperature by cycling water slowly through the flat panels until it warmed enough, then into the tank at whatever rate is needed to keep the temperature constant. Then the water feeding your downstream heater from the tank can be at a consistent temperature. I don’t know if anybody sells that kind of thing already built, but it seems like a couple of positive-pressure water pumps, a microcontroller, and three or four thermistors ought to be the entire materials cost.

  6. …oh, I’m an idiot, I didn’t notice that the very first paragraph of the post explains that it was a 750-gallon tank (2800ℓ). That’s 14 200ℓ drums’ worth, which would be US$350 to US$1400 at the new prices I cited above, or US$112 used from your local Coca-Cola bottler, if the anecdote I found earlier is to be believed. Screw PVC pipe into the tops with some PTFE tape, glue on elbows to route from one to the next, set ’em on straw bales, and you can maybe get the whole insulated tank complex built for under US$500, surely under US$1000. Won’t rust, won’t leach, and I think that without sunlight, it won’t degrade in 5 years or 20 years. With livestock tanks you might be able to get it even cheaper.

    If you wanted the barrels to be able to be partly full, as in my earlier suggestion with the thermostat-controlled water pump, you’d need the water to go in and out of the bottom of the barrels instead of the top, and you’d need a way to let air in and out the top, and then prevent overflow with a float or two (because it wastes energy and water) and channel overflow with, say, a big tray (because you don’t want it to also waste your straw bales).

    Also occurred to me that the 125psi of your US$4350 tank might be from the water main rather than unwanted boiling. The typical way to cope with unwanted pressure here (say, from the tank on top of a 20-story building) is that the water tank somewhere high up (say, your town’s water tower) feeds a tank lower down, open on the top, which is controlled by a float valve, like a toilet. When the float is low, the valve opens and water pours into the lower tank; the energy from the extra pressure is dissipated in turbulence. When the lower tank gets full, the float lifts and the valve closes. You put the lower tank high up enough that it gives you the pressure you need, but not so high that you have to buy expensive high-pressure tanks, and with a drain nearby so that the spilled water won’t destroy anything expensive when the float gets stuck.

    I can’t imagine you need much pressure for baseboard heating. Even if you do, you could probably achieve it with an electric pump placed somewhere after your 2800ℓ of warm water, so you don’t need a high-pressure stainless-steel warm-water tank that can withstand the pressure.

Leave a Reply