How much power does it take to turn an alternator
 

This gives you an idea of how much:
Mopar Engines - Power Vs. Luxury - Mopar Muscle Magazine
(found originally by t vargo)
The hot rod article says 3 to 4.5hp depending on how fast you spin the very small 60 amp alternator they used. Its reasonable to assume the far larger 200 amp AD244 I am going to use will take even more engine power to free spin.

Going to run a test that might answer this question. A small scale test before a larger build.
For the test engine I have my brand new 5.5hp Honda horizontal shaft engine I bought to make a portable air compressor out of.
The alternator will be an AD244 with the voltage regulator replaced with a big rheostat.
The idea is to take the engine, run it at various no load speeds to see how long a set amount of gas runs for. Then put the alternator with no load on the engine and see how long it takes to burn through the same amount of gas.
We will assume 0.5lb/hr per horsepower.
Then redo some of the tests with a load. Double check some to ensure consistency.
The main objective is to calculate horsepower to turn the alternator at max output for that engine with various pulley sizes (2.6'' on the alternator and 4 and 5 inch on the engine), that should give me about a 1 to 1.5 ratio and 1 to 2 ratio. The max power tests will likely need to do the loaded test first and then run the engine that speed with the alternator running no load after. Max power tests should be run at about 5,000 alternator RPM and about 6,800 alternator RPMs.
Then later on scale it up to see what it would take to run this alternator full field output at target speed, some where around 6,000 or 7,000 RPM.

Twenty years ago I maintained a small fleet of ambulances for our city of 7000 people. One thing that I noticed was the belt tension and belt life on a single 200 amp Leese-Neville alternator. I had to replace the single 1/2" drive belt with an OEM belt every other oil change to have a reliable ambulance.
You only loose one belt on an ambulance before you step up your game!
The situation got much better when the manufacturer went to twin 125 amp alternators with a 6 rib belt.

I am going to use industrial sized half inch belts and I still don't think the belts will last a real long time with what I have planned for them.
For testing I believe I will be using an automotive 1/2 inch belt.

This spells it out pretty clearly.
Note this is a 24v alternator.
http://ecomodder.com/forum/attachmen...1&d=1469680926

When I am driving, if at a steady load acceleration with a somewhat depleted battery I switch on my alternator I can feel the load engage. It shudders the car in that same subtle way you sometimes feel DFCO or an air conditioning compressor engage.

Wait, why are you testing this? You can find alternator efficiency charts everywhere. 60A load * 14V / 50% efficiency = 2.25hp, that would be a pretty typical load I think.

Calculating the amp draw power is easy. I am more interested in the now load power draw since we will assume a fairly typical ecomodder car will use very little power.
Seems everyone that makes these charts focuses on how much brake horsepower these use when making big power. We don't need big power.

A chart like this?
http://ecomodder.com/forum/attachmen...1&d=1470048377

So I took a heavy duty pickup and SUV alternator and hooked it up to a 5hp Honda GX200 engine.
One thing I found was that energizing the exciter field even with no output load put nice little load on the engine. I was not expecting that.

http://ecomodder.com/forum/attachmen...1&d=1471620551

Eddy current and iron hysteresis losses.

Alternator laminations are usually on the thicker side for cost savings, so they have higher losses.