In terms of viscosity, this Shell fluid almost looks like sludge at 0 C (32 F). Also the heat capacity and thermal conductivity are low (compared to water).

I would design this system to prevent freezing (-30 C, as you specified) using an alcohol/water mixture. Minimize the distillation issue with a low percentage glycol that shifts the fraction (distillation) point higher in temp.

Am pulling physical chemistry labs out of my brain from more decades ago that I prefer to mention.

The major issue is freezing. Due to the low operating temp, an organic fluid appears way to viscous in the range of temps you operate in.

You also have huge summer insolation rates and you may want to put in a back-up circulation pump to prevent summer heat stagnation and a vapor expansive calamity (if you use alcohols).

VERY interesting problem.

Steve

At temperatures below ~25C (77F), there wouldn't need to be any pumping or heat transfer, just the ability not to break anything, as freezing of water would do. Sludge is fine at 0C, as long as it becomes less viscous at normal temperatures again.

When the circuit is quite small, how would that affect it for having less heat capacity? Isn't it a question of pumping slightly faster, or is the difference too big between water/glycol and these types of oils? All it needs to do is go through the panels then just inside the roof to a heat exchanger - from there, it's all water in the rest of the system.

If I did use a glycol/water liquid at something like -30C freezing point, I could design a reverse cycle into the system for those very few days it should get below -30C, or even close to it - just to be safe. I know it's counter-productive, but it wouldn't take much to cycle the liquid a few times during a cold winter night - just to keep it from freezing up.

As for the lower boiling point of either water or water/glycol - what would happen if it really stagnates and boils? I would imagine it would blow the overpressure valve eventually, unless I have a disproportionally huge expansion tank.

I do have a DC direct drive pump built into the design, both on the primary and secondary side of the heat exchanger. If power is available from the PV panels, and the temperature exceeds a set limit, these pumps will start, regardless of the controller and the rest of the system.

I am trying to post a diagram of the setup design overview, as it is right now, but it was reduced to quite a small size, so I am trying again, in a different way.

Some text has been cropped in the diagram, as I have problems exporting from TAPPS (Windows XP!) to something readable.

DC1 and DC2 pumps will run if 1) electricity is received by PV panels and 2) S1 temperature probe is exceeding a set value (possible 75C to 80C, depends on liquid and certain other items). That way, those pumps will augment the existing pumps in case of AC power failure, controller failure, AC pump failure or other situation causing overheat condition.

As I am not always at the house to fix it, I would prefer some liquid which would not boil away in case of a failure, hence the oil mentioned previously sounds interesting (despite viscosity and heat capacity problems).

A liquid that never boils nor freezes, at any expected temperature whether there is system failure or not, is quite tempting to consider. Especially if it lasts a long time, and doesn't gel up.

https://dl.dropboxusercontent.com/u/18397/Diagram.PNG

The problem with what you propose is this: what happens during the winter when the sun shines bright? The oil in the collector heats up and thins out, while the plate hx is clogged with tar. The thin oil sheds its heat to warm the cold oil before it can effectively move fast enough through the loop to warm the house.

If you really need that high of a stagnation temperature, the oil will expand, too. Not as much as boiling water, but enough that you will need a holding tank. Something open to atmosphere.

The new kid on the block is BETAINE (BEET-uh-een), aka trimethylglycine. It is derived from sugar beets as a by-product of, you guessed it, the sugar manufacturing process. Being the entrepreneurs that they are, the sugar beet farmers have improvised a new product from their waste.

Here's a product flyer:
http://www.climalife.dehon.com/uploa...-uk-08-pdf.pdf

The only problem I see in your application for this product is its high temperature limit. Above 100 degC, it starts to disintegrate/decompose.

Biproduct from sugar, sweet!

I see this is a European product, should make it even better for me. I saw that company also has another product which does boil around 105C, but it's stable up to 150 (without gelling up)

As I wrote before, I do intend to keep the system cooled off all the time, both for efficiency and to avoid overheating of PVs (it's a hybrid system) and the heat transfer liquid.

The only concern is what happens if - despite redundancies - the panels do overheat.

The heat exchanger is indoors and should be fine, but the issue you mention is valid, mainly for the content in the tubes between the panels and the indoor part.

Not very many days a year are actually below -10C, and even more seldom -20C. I have been considering a very slow running of the system in case of severe cold, just as an extra precaution. Maybe a one-minute flush when it reaches a certain temperature. Since it's not very often, the loss of energy could be a small price to pay for "extra insurance". Nothing as annoying as busted pipes or panels!

Looking at your diagram, it seems to me you could put a condenser in pretty easy as a heat dump. It's real easy, and I'll quickly tell you now, a step-by-step procedure.

1. Pick a common midsize car. For me, this would be a toyota camry or a ford taurus. In europe, maybe bmw 3-series?

2. Go to junkyard and get a condenser with the electric fan(s) attached...they call it a radiator with fans. If you can, get the expansion tank also.

3. Plumb the condenser in place of the overpressure valve in your system. Install two tees in the loop. Orient both tees vertically with the center fittings facing horizontally, one above the other. The center fitting of the top tee should be connected to the outlet of your collector. The center fitting of the bottom tee connects to the lower radiator hose, the upper radiator hose connects to the upwards facing fitting on the top tee. The downwards facing on the lower tee connects to the plumbing going to the main heat exchanger.

4. If you also obtained the expansion tank, make it the high point of the outdoor loop. The filler cap will serve as the overpressure valve.

5. The condenser should be positioned so that it is normally mainly empty, and fills only during a distillation or boiling event. In case of an inferno, the electric fans can be activated via staged thermostats. The latent heat transfer capacity should be well within your 10kw maximum.

With this rig, the condenser will regulate the max loop temperature to that of the boiling point of the most volatile fraction of your heat transfer solution. I would recommend a methanol or ethanol mixture, as the distillation would begin around 90 degrees Celsius at atmospheric pressure. Maybe closer to 100 under some pressure.

If the condenser/fan unit can keep up with your generated heat, you will retain all your coolant. If your main heat exchanger has fully charged your heat store, and/or cannot keep up with your generated heat, the condenser radiator can be run either passively as a condenser or actively by adding a pump between somewhere and the upper radiator hose.

If the temperature rises enough, the alcohol will begin to separate from solution, carrying its latent heat at high volume to the condenser. As the alcohol boils off, the boiling point will rise, providing a small buffer. Adding an electrolyte to the solution will increase this effect due to the increased boiling point of the saltwater brine.

The coolant will not all boil out of this system. At a certain level, there will not be enough liquid in the loop to either heat sufficiently to boil or to pump constantly, depending on the plumbing configuration. At this point, your main heat exchanger will still be submerged. If an electrolyte was added, it will be concentrated in the remaining water and will serve as freeze protection.

Thanks, Jeff, I will run this with my plumber, see what he says. I didn't get yet exactly how it works yet I can imagine how it must work. I'll make a drawing, step by step, from what you wrote, then I am sure I'll get it.

Thanks again, for taking your time for this.

Jeff,

This is BRILLIANT! I had not thought of using a junk auto radiator. But the entire thing is neat - especially the expansion tank and filler cap "pop off" valve.


Thanks,

Steve