VidPro,
The solar grid tie inverter I will have installed is around 94% efficient--not bad at all (150 watts of loss on 3kW of power). It is hardly worth trying to capture that waste heat on a cold winter day.
Xantrex 3kW Grid Tied Inverter
Your comments on keeping solar radiation from entering the house in the first place are absolutely correct. Adding shade (trees, awnings, etc.) are also good options. Making sure that you use CFL's and keep overall power use low helps too (such as venting your stove/kitchen and/or only cooking big meals when it is cooler to keep heat out of the living area during the hot part of the day, etc.).
Regarding matching the sine waves to the utility... Technically, not matching the sine wave current gives you a poor power factor but an inverter will work OK
In newer computer/electronic equipment power supplies (over ~100 watts or so--don't recall exactly) having Power Factor Corrected supplies is mandatory for many countries. And is just good sense because of I^2R losses associated with the higher current of non-linear and poor Poor Factor equipment (for non-technical folks, think of hooking a big Cap or Inductor directly to an AC outlet--basically the utility is supplying current but can't charge for the power because there is no work being done--current is +90 or -90 degrees out of phase with voltage and no work is done--except for I^2R heating losses--everything works fine, but the utility supplies current without being able to charge for its full value, i.e., work--big industrial sites, like refineries, will pay penalties for power power factors--utilities have capacitor banks that they can switch in and out to handle AC and irrigation loads during the summer).
All that said—it is still not safe or wise to connect a home made inverter to a utility even if it was designed well. The other part of the "contract" is that the company goes through regulatory testing and ongoing manufacturing safety controls (with companies such as UL—Underwriter's Laboratories) to ensure continuing safety and compliance with the requirements.
<font color="red">Disclaimer: Below is a suggested way of doing a cheap and dirty grid tied power. I have used elements of this before without any problems—but I have not done this as a grid tied power system. And therefore, the following is placed on this board for discussion and experimentation purposes only. As with all grid tied systems, there are inherent dangers and unknown effects and consequences that are not addressed by the following. Anyone wishing to experiment with the following system is entirely responsible for the legality and safety of their system. </font>
All that being said—there is one sneaky (and, I believe relatively safe and legal) way of putting power back into the AC mains. This would be the Motor/Generator set. Basically get a DC motor sized for your solar array and connect it to
any normal AC induction motor (not synchronous or AC/DC brush motor or true AC generator) and connect the AC induction motor to your AC lines through a relay that is controlled by your solar panel output (or some other sensor to make sure that you are generating positive energy and not just spinning everything from the AC Line).
Basically, the way this works is that your induction motor starts and will run the motor/gen set at ~1,725 RPM or so (consuming AC power). Now, power up the DC motor with the solar cells and start driving the induction motor at over 1,800 rpm (probably at 1,875 rpm or what ever the inverse of the name plate of the induction motor is).
This works because there is a magnetic field in the armature of the AC Induction Motor induced by the rotating magnetic field of the external stator that was setup by the AC mains. And it is safe because if the AC mains fail, there is no current/voltage to maintain the fields and the output will drop to zero until the AC mains are turned on again (this is not an AC generator and there is no "self exciting" of the "generator" for it to generate power in the absence of AC line power. You would also need to understand the DC side of the circuit and make sure that any DC voltage generated by motor/gen set does not damage your solar cells.
The down sides of the above—probably terribly inefficient (probably in the 50% range) and I would want to experiment on making sure that the system behaves safely in an AC Mains brownout/power loss situation. Also you would have to make sure that the control circuit is setup so that it only runs when positive work is being done by the solar (monitoring RPM is greater than 1,800 RPM, or output current, or estimated based on Solar Power input). The advantage is that this system would work for any home made power source (wind, hydro, bio-fueled generator set) without costly electronics and can be built from scrap yard components.
That all being said, this is based on how I have generated 3 phase AC power for large mills and lathes in a business that only had single phase power available (you hook up a 5 HP 3 phase Y 3,450 RPM wound motor to your single phase power, give it a spin with a small motor to get it rotating, then hit is with the single phase AC power. You connect this 3 phase Idler motor to the reset of the 3 phase equipment in your shop… Actually works very well. I am sure that the generated phases are not great, but you could not tell the difference when running equipment up to ~2.5 HP. You need to make sure that any control circuits are connected to the "true" single phase power and not to the generated third leg.
-Bill