PASSIVE SOLAR ADAPTATIONS IN EXISTING HOUSES, PART 2

We left off with a minor passive solar project outlined, a 1000 square foot bungalow winterized and solarised in Columbus, Ohio. Calculations for a clear day in January suggested that about two/thirds of the heating load for the day would be provided  by passive solar. The additional  investment to achieve this result would be minimal. A contrary example, with the same footprint, but wrongly oriented glass of equal area, performed poorly in comparison.

We did make a few assumptions; firstly, the house has good solar exposure in winter. This is non-negotiable, clearly. The second assumption is that the house has a long wall facing south, i.e. the long axis of the house is more or less on an east-west line. This is much more negotiable. Even an orientation 30 degrees away from the ideal or a different plan shape is still workable.

We made  a third assumption of a ground floor slab-on-grade with 3 inches of under-slab foam insulation. This is unlikely to be the case. More often, the average house in a cool climate will have a basement with an uninsulated slab that gets almost no sun, and with a joist and strand-board and carpeted main floor over. The foundation walls  are masonry or concrete, but  probably insulated on the inside. All this mass in the basement is useless for heat storage.

The available thermal mass of the house will be much smaller, roughly 60 cubic feet of the gypsum core of the drywall on the ground floor being the only significant component with a heat capacity per cubic foot close to that of concrete, but somewhat insulated by the thick paper facing. We'll use this more typical bungalow with basement as our example as we proceed with our investigation.

We hinted in the first part of this article that an economical and elegant means could be found to achieve the effect of thermal mass without attempting to retrofit the house with an impractically bulky and massive thirty-six ton Trombe wall.

John Michael Greer, in his post Alternatives to Absurdity characterized the passive solar techniques of the 70's - including thermosiphon air panels, Trombe walls, and attached greenhouses, as baby steps towards learning to live comfortably on nature's diffuse energy flows. I'd like to suggest an order of magnitude improvement to the design problem of heat storage in a passive solar house, an idea of tremendous potential that was indeed researched, patented, but with the subsequent cheap oil and good times starting in the 80's, was never commercialized, and remains largely overlooked. A curious fact, that such an energy-saving invention was buried.

Consult a patent attorney on this point if you're keen on going into a PCM business, but I'm fairly sure these old patents have expired, and are now in the public domain. You'll be searching for reversible phase change materials; remember this clue. Allow me to explain.

A Trombe wall, like a floor slab, stores sensible heat, so called, because the mass simply warms up as it absorbs heat, and cools down as it loses it, as anyone knows. We can sense the energy in the mass by touch, obviously, or even by proximity, hence the term “sensible”. 

It’s always necessary to get a grasp of quantities, however. We had sized the Trombe wall according to The Passive Solar Energy Book. If we assume that a temperature range of 65°F to 75°F is the most that would be acceptable to the occupants, then we can calculate the amount of heat the wall can store, using Equation 4, Q=∆T•V• H_cap.

Q= 10°F x 480cf x 22.5 BTU/cf-°F = 108,000 BTU.

We signed off our last post, you will recall, with a question "What is the ice doing at that moment?", referring to ice cubes in a glass of Scotch our thermal physicist was enjoying.

Obviously, the ice was melting and cooling the Scotch. Our physicist might elaborate, and reveal that the ice, cooler than the Scotch, was absorbing thermal energy from the Scotch and undergoing a change of phase from solid to liquid, at a constant temperature of 0°C, with an enthalpy of fusion of 334 kilojoules per kilogram, which equates to 144 BTU per pound.

The enthalpy of evaporation of water is about seven times larger. Wonderful stuff, water. The physical and chemical properties of water, and its sheer abundance, makes life possible on earth.

Other terms for enthalpy of fusion are latent heat or heat of fusion (there is no change in temperature, but energy movement associated with a change of state), as distinguished from sensible heat. A quantitative comparison of latent heat with sensible heat is in order just now.

Equation 5: Q_fusion = M•H_fusion   where M= V•δ, i.e. mass equals volume times density

Let's figure out how many cubic feet of ice 108,000 BTU could melt. Ice has a density of 57.22 lbs/cf.  Solving Equation 5 for volume,108,000/(57.22 x 144) = 13.1 cubic feet, about the size of a dishwasher,  with a mass of 750 pounds. Compare this to our notional Trombe wall at 480 cubic feet and 72,000 pounds, that would use up one sixth of the floor area, and have to warm up by 10 degrees Fahrenheit, to store the same thermal energy.

This illustrates that a latent heat storage strategy trumps a sensible heat storage strategy, such as a Trombe wall. These are greater than order of magnitude differences, exactly what is needed to expand the feasibility of passive solar space heating  from the sunny South-West  into the cloudier, yet populous, temperate zones of the continent. A further and considerable advantage of latent heat storage is a stable temperature within the storage medium, eliminating the overheating or control problem we found in Part1, (discussed by Mazria on page 65 of my copy) and thus offering the promise of temperature control within the living space.  

Of course, humans aren't polar bears, we prefer an environment at around 70° F, not at freezing, so ice won't work. What we could really use is a material that melts at or a little above room temperature, with a substantial enthalpy, that isn't dangerous, and is inexpensive, obtainable and will keep cycling (storing and releasing heat) for a good many years.

Before we leap into this new area, let's just review our progress for a moment. In Part 1, we looked at conductive heat loss for the house. We examined infiltration and ventilation losses. We added our occupancy gains from warm bodies and electrical use. We then looked at our expected solar gains on a sunny day with 120 square feet of windows on the south wall, and saw that it generated two-thirds of the heat needed for a January day and night of average temperature. We then compared these results with a contrary house with the windows on the north side, which performed poorly. 

We'll next consider days with only average sun, we'll add insulating night blinds, and more south windows. Since we will have a warm thermal mass in the house, increasing the average temperature of the radiant field that is the house interior, let's make an educated guess that if our thermostat is around 68°F (our indoor air temperature), we'll still be comfortable. See The Passive Solar Energy Book, page 64, for a brief discussion  and chart on the topic. We can deduct another 2°F  as a night setback.  We're moving closer towards a passive solar house.

If we look at insulation and infiltration standards in a Passivhaus, the European system getting a bit of notice on this side of the pond, and translate into our units, we see recommendations including about R 70 attics, R 50 walls, R 30 foundations and slabs, R 6.6 windows, and an ACH of 0.6. We'll try to approach these values in our retrofit, but with our eyes and minds open.  

Increasing our attic insulation with a bit more inexpensive cellulose (recycled and treated newspaper) insulation while we're doing that job anyway, is fairly easily accomplished.

As for the wall R-value, again, since re-cladding and an air barrier is part of the project, why not use an insulating sheathing with an integral low-e air barrier facing? This can add another R 11 to the wall value at moderate cost. Being continuous over the wall surface, it will also minimize the thermal bridging effect of studs, sills, etc. of the framed wall.

The foundation approach will vary with the conditions inside and outside of the wall - are there existing interior finishes? Is weeping tile to be installed? Then maybe the insulation goes to the outside. Most often, it is less expensive to work from the inside, so let's assume that's the case. It's important to ensure that the critical floor to foundation juncture is properly and well sealed against vapour and air movement.(Vapour barrier on the warm side, air barrier to the outside.)  Often, local codes only require the insulation extend two feet or so below grade. We'll assume we insulate full height, with R 13 batt, R 10 open cell foam insulation/vapour barrier over that, and a drywall finish. We won't attempt any further breakdown of the heat flows due to Delta T's varying with the depth of the foundation; those who like a challenge can consider this elaboration however.

The basement slab can be insulated with 3" of closed cell foam for R 15, and finished with a click flooring system, though local codes will likely require a fire barrier over the foam. Another point to consider is how this added dimension will affect headroom and the stair to the basement. All in all, the slab, given the small Delta T for this element, is probably an item that will be dropped if budget considerations come into play.

One adjustment to the spreadsheet that the homeowner can accomplish at no cost, is damping down the heating vents to the basement. Let's adjust the basement slab and foundation Delta T down by 2° F. Maybe an even cooler basement will be preferred if the homeowner gets serious about home brewing - that's an entirely different, but interesting topic.

If we recalculate the wall, foundation and slab R values allowing for air spaces and films, and all materials in the assemblies, per ASHRAE heat transmission coefficients, we get values of 35.5, 25.7, and 16.3, respectively, and respectably.

As for windows,  a whole window value of R 6.6 is achievable only with triple glazing, two or more low-e surfaces, argon or krypton fill, insulating spacers, and insulated frames; in other words, high-performance, premium windows. The SHGC of the glazing will be around 0.5, versus the 0.69  value we used in Part 1, i.e. we'll only admit about 5/7 of the solar gain of double-glazed, low-e windows.

 In terms of energy efficiency then (cost aside), triple glazed premium windows are of benefit only if we do not use insulating night blinds. ( I did the spreadsheet runs, it is so.) The blinds then, may well be the more attractive and economical option. They need to perform properly, of course, with a snug fit and a vapour retarding inner face to minimize condensation on the glass. We'll assume we can add R 8 for the night blinds, assuming they're pretty good. The fairly standard new low-e and argon filled double glazed windows I just had installed last season have a whole window value of R 3.3, so we'll use this figure in our next spreadsheets.

A final tweak: we'll insert a diurnal variation of plus or minus 4°F around the average outside temperature, distinguishing night from day temperatures, as those insulating blinds make a difference at night that we should take into account.

  CHART 4: AVERAGE SUN & TEMPERATURE WITH STRATEGIC IMPROVEMENTS

HEAT LOSS/SOLAR GAIN CALCULATION, IMPERIAL VALUES						HEAT GAIN  SOUTH GLAZING
40 N LATITUDE, COLOMBUS, OHIO						INSULATING NIGHT BLINDS R			8
JANUARY AVERAGE TEMPERATURE, F				32.1		NIGHT SETBACK			-2
AVERAGE DAY INSOLATION						INDOOR TEMP  F			67.6
DIURNAL VARI'N  4
DAYTIME CALCULATION			12	HOUR	DELTA  T	31.5	SLAB/FDN DELTA T		11
COMPONENT		AREA	R-VALUE		BTU/HOUR		DAYTIME TOTAL BTU
WALLS		840	35.5		-745			-8944
ROOF		1000	70		-450			-5400
N/E/W WINDOWS/DR		60	3.3		-573			-6873
SLAB		1000	16.3		-675			-8098
FOUNDATION		1040	25.7		-445			-5342
SOLAR WINDOWS DAY		240	3.3		-2291			-27491
SUBTOTAL								-62148
			GAIN	FRAME	SOLAR GAIN
			DBL GL	ALLOW	COEFF LOW-E
NORTH WINDOWS		20	120	0.8	0.92			1766
EAST WINDOWS		20	240	0.8	0.92			3533
SOLAR WINDOW GAIN		240	603	0.8	0.92			106514
DAYTIME SOLAR GAIN								111813
NET DAYTIME GAIN								49665
NIGHTTIME CALCULATION			12	HOUR	DELTA T	37.5	SLAB/FDN DELTA T		9
COMPONENT		AREA	R-VALUE		BTU/HOUR		NIGHTTIME TOTAL BTU
WALLS		840	35.50		-887			-10648
ROOF		1000	70		-536			-6429
N/E/W WINDOWS/DR		60	11.3		-199			-2389
SLAB		1000	16.3		-552			-6626
FOUNDATION		1040	25.7		-364			-4370
SOLAR WINDOW NIGHT		240	11.3		-796			-9558
SUBTOTAL								-40020
WHOLE BUILDING NET 24 HOURS CONDUCTIVE + SOLAR GAIN/LOSS								9646
			VOLUME	ACH	H-CAP AIR		AV DELTA T
INFILTRATION LOSSES/DAY			8000	0.5	0.018		34.5	-59616
VENTILATION LOSSES/DAY			INCLUDED
			KWH		BTU/KWH
DEGRADED ELECT'Y CONTRIB'N			12	0.9	3412			36850
OCCUPANCY CONTRIBUT'N/DAY			3	12	400			14400
WHOLE BUILDING NET 24 HOURS GAIN/LOSS ALL SOURCES								1280

  CHART 5: SUNNY DAY & AVERAGE TEMPERATURE WITH STRATEGIC IMPROVEMENTS

HEAT LOSS/SOLAR GAIN CALCULATION, IMPERIAL VALUES						HEAT GAIN  SOUTH GLAZING
40 N LATITUDE, COLOMBUS, OHIO						INSULATING NIGHT BLINDS R			8
JANUARY AVERAGE TEMPERATURE, F				32.1		NIGHT SETBACK			-2
CLEAR  DAY INSOLATION						INDOOR TEMP F			67.6
DIURNAL VARI'N 4
DAYTIME CALCULATION			12	HOUR	DELTA  T	31.5	SLAB/FDN DELTA T		11
COMPONENT		AREA	R-VALUE		BTU/HOUR		DAYTIME TOTAL BTU
WALLS		840	35.5		-745			-8944
ROOF		1000	70		-450			-5400
N/E/W WINDOWS/DR		60	3.3		-573			-6873
SLAB		1000	16.3		-675			-8098
FOUNDATION		1040	25.7		-445			-5342
SOLAR WINDOWS DAY		240	3.3		-2291			-27491
SUBTOTAL								-62148
			GAIN	FRAME	SOLAR GAIN
			DBL GL	ALLOW	COEFF LOW-E
NORTH WINDOWS		20	120	0.8	0.86			1651
EAST WINDOWS		20	474	0.8	0.86			6522
SOLAR WINDOW GAIN		240	1506	0.8	0.86			248671
DAYTIME SOLAR GAIN								256844
NET DAYTIME GAIN								194697
NIGHTTIME CALCULATION			12	HOUR	DELTA T	37.5	SLAB/FDN DELTA T		9
COMPONENT		AREA	R-VALUE		BTU/HOUR		NIGHTTIME TOTAL BTU
WALLS		840	35.5		-887			-10648
ROOF		1000	70		-536			-6429
N/E/W WINDOWS/DR		60	11.3		-199			-2389
SLAB		1000	16.3		-552			-6626
FOUNDATION		1040	25.7		-364			-4370
SOLAR WINDOW NIGHT		240	11.3		-796			-9558
SUBTOTAL								-40020
WHOLE BUILDING NET 24 HOURS CONDUCTIVE + SOLAR GAIN/LOSS								154677
			VOLUME	ACH	H-CAP AIR		AV DELTA T
INFILTRATION LOSSES/DAY			8000	0.5	0.018		34.5	-59616
VENTILATION LOSSES/DAY			INCLUDED
			KWH		BTU/KWH
DEGRADED ELECT'Y CONTRIB'N			12	0.9	3412			36850
OCCUPANCY CONTRIBUT'N/DAY			3	12	400			14400
WHOLE BUILDING NET 24 HOURS GAIN/LOSS ALL SOURCES								146311

The observant reader will notice a factor of 0.92 in Chart 4. This appears because we have used Mazria's Appendix 3, Average Daily Solar Radiation, see pages 348 and 358, to generate the 603 BTU/sf-day figure for south double glass. This is adjusted by 0.92, the ratio of our low-e glazing SHGC of 0.69 to the 0.75 of double clear glass. The north and east values are plugs - best guesses - as Mazria offers no figures, but any errors are fairly insignificant.

On the average day, our bottom line is now actually a small gain, thanks to the adjustments made. We are essentially balanced for passive solar. Looking at the new bottom line for a sunny day, we see we're running a large surplus, or net gain for the 24 hours, in the coldest month of the year. We've made great progress towards personal energy freedom and reducing the household carbon footprint. The longer days and gentler temperatures of the rest of winter will show even better results

Recall that the indifferently oriented second example in Part 1 managed to lose100,000 BTU on a sunny January day. It's possible, in other words, with a bit of planning and effort, to go from a loser to a winner in solar terms, and harvest the sun with the simplest thermodynamic system conceivable. It works without additional machinery, as we'll see!

From the six hours of peak sun on our clear January day, we have a daytime surplus of about 195,000 BTU.  If we sized our latent heat storage to handle this and hold say the net surplus of 146, 000 BTU from the preceding sunny day, we'd need about 340,000 BTU of storage.

Since our phase change material actually exists, with a melting enthalpy about half of that of ice, at 75 BTU/pound, a density 50% greater than water, and a melting point that can be actually designed to fall  within a range just above room temperature, and can go through 10,000 freeze/thaw cycles with only minimal loss of performance, we're almost in business. And importantly, for entrepreneurs everywhere, the US patent is long expired.

How much material do we need? Do the math, and 340,000/75 comes out to 4500 pounds. At a density of about 94 pounds per cubic foot, this is just about 48 cubic feet of the stuff, almost 1400 litres. We have to think about heat flow into and out of the material, so it can't be just a big lump in the room. The material has a reasonable thermal conductivity, so cylinders of 2" or so diameter might work all right, say recycled 1/2 litre soda bottles, or mineral water bottles, which have an attractive appearance like miniature wine bottles. And hey, we're reusing consumer waste to save fossil fuels. You've got to love that!

However these containers are arrayed, we'll need about 2800 of them! But if they stack up at say 25 per square foot on their side, that's 112 square feet of face area. The simplest, most direct placement of our phase change material array, or PCM array, is directly in front of the south windows, with a foot or so of space to allow access to our insulating blinds. Our south windows probably take up about 36 linear feet of the wall, so we could easily fit our array, complete with any cabinetry or metalwork frames and tops, into the full window length and end up at about a three foot height. Other arrangements, of course, are possible, should we want to go for broke and come closer to the full height of the windows to intercept as much sun as possible. The containers should have a dark, heat-absorbing color, but dark greens, reds, blues or browns are workable alternatives to black. The colour could be as dyes in the PCM, if the chemistry wasn't adversely affected, or as surface color on the containers. As well, the container surfaces should have average emissivity, which they do as found, though many folks will want to add a decorative touch. Avoid metallic paints, however, or paints with nasty, plastic-dissolving solvents. A job for the artist in the family.

As for carpentry, that additional dead load of about 125 pounds per linear foot for the PCM close to the south foundation is now well within the capacity of the wood floor framing to support. It's best to check this against actual conditions however, and make any reinforcements necessary.

We would benefit from having some ceiling fans to gently move the air around to improve the charging (melting) of the PCM on a sunny day, and to assist discharging (solidifying) in cooler weather. They are great for summer comfort as well, and are likely already in place.

The potential for this strategy seems pretty attractive, doesn't it? Pound for pound, the phase change material has about 50 times the storage capacity of the concrete Trombe wall (which has a 10°F operating variance), yet with the advantage of a constant temperature.

Obviously, some fine tuning may be in order. For instance, we can infer from the above charts that the nightly heat loss of the house, is 40,000 BTU, or 3300 BTUH. Ideally, the passive discharge loss for the PCM array, i.e. without any fans running, should be no more than 3300 BTUH. This will be a function of the inside air temperature, the PCM melting temperature, the geometry of the array and its enclosing structure, the emissivity of the containers, and the thermal resistances of the containers and of  the PCM itself. None of which is much of a challenge for a competent thermal engineer or even a committed layman. Look up ASHRAE Fundamentals to start on this. I expect, though, that our recycled pop bottles will be in the zone.

We also can infer that on the morning after that second sunny day, we will have about 300,000 BTU still in storage. This is enough to offset  a few days of bad weather, as our average days are no net thermal loss. In all, we'll knock off a high fraction of our winter heating load. An hourly simulation could give us an estimate to the nearest percentage, but I haven't hired that engineer yet.

Even single digits reduction of national energy use is a boon in our predicament of rising energy costs, and impending shortages, particularly so as shelter is fundamental to human health and well-being. Diffuse solar radiation for low temperature space heat. It doesn't seem all that far out, does it?

But I have to ask, why is the most economical bulk commercial source of the phase change material in Mumbai, India?  The last time I checked, before the recent financial excitement, PCM P. Energy would ship by the ton to a North American port for 1.25 USD per kilo. Kind of dumb, isn't it, importing a composition made up mostly of road salt from India? I'm pretty sure we have calcium chloride here, and chemists, engineers, and entrepreneurs. Get with it, America!