PTO Generator

[TugboatEng]

I have an idea that requires a generator that can produce 10-20kW of power over an available speed range of 650-1800 rpm, it's essentially the same application as an automotive alternator. I can use a step up gearbox if necessary. I eventually need it to be 480V 3 phase. I'm looking for guidance with regards to the best type of generator to achieve this output. Bonus if it can be motorized.

 

[waross]

I think it's easier to modify the speed than deal with variable frequency or voltages.

With DC, frequency doesn't matter.
Variable voltage?
An AVR will take care of that.
I have seen diesel generators where the voltage will vary from 100% at no load to 50% at full load with fixed excitation.
Add an AVR and the voltage will be stable within a few percent.
Why do aircraft use 400 Hz?
Size and weight.
A example 10 Volt transformer may supply a maximum load of 1 KVA at 60 Hz.
That same transformer will supply a maximum load of (400Hz/60Hz = 6.67) 6.67 Hz at 66.7 Volts and 400 Hz.
This is based an a safe Volts per Hertz Ratio. (1:10 The voltage in Volts shall not exceed 1/10 the frequency in Hz.)

What to expect in operation?
This example is based on the internal voltage drop characteristics of a conventional diesel generator.
The speed ratio from no load to full load capability may be less but the principle is the same.

With fixed excitation, the voltage of an alternator is speed dependant.
In this example, we will use a target voltage of 50 Volts from a 24 Volt automotive alternator.
In our example this alternator no load output voltage will reach 24 Volts Nominal/27.6 Volts actual at an engine speed of 550 RPM.
Using the 2:1 ratio, the alternator will reach 24 Volts nominal at 1100 RPM.
At speeds above 1100 RPM, the AVR will hold the voltage steady at 24 Volts and the available output power will no longer increase with increasing speed.
What happens at 825 RPM or 50% of the speed range?
The alternator will put out 50% of rated current before the internal voltage drop causes the terminal voltage to drop.
The actual current may depend on the state of charge of the batteries.

Now what happens if we want to run this alternator at two times rated voltage?
ALL of the speeds given in the example will be doubled also.

The other parameter to consider is the minimum frequency.
The Volts/Hertz ratio (to be determined by you) must not be exceeded.
You may have to add a circuit to inhibit excitation when the V/Hz ratio is too low.

More on voltage step-up and transformer sizing later.

 

[FacEngrPE]

Caution - Making transformers that work well over a range of frequencies (650 to 1800 RPM gives a output frequency ratio of 3.2:1). The operating frequency impacts the number of winding turns needed, the amount of iron, winding impedance, saturation point etc. I doubt this is an off the shelf transformer.

A lower cost solution can likely be developed using an off the shelf alternator with like the prestolite units waross suggests. The literature is a bit thin as posted here https://www.idleproextreme.com/ so you likely need to discuss with the factory to get details on the included regulator. This solution is limited to 24 volt battery systems, if you need higher voltages DC DC Boost converters are available.

On DigiKey's website DC/DC BOOK OF KNOWLEDGE Practical tips for the User By Steve Roberts M.Sc. B.Sc., Technical Director, RECOM

 

[TugboatEng]

DC to DC converters are expected to be involved . 480V 3-phase 60hz is ideal because it would be the most efficient for normal operating and the excess available power could be DC boosted for charging. However, straight DC output for charging and then inverting from the battery is another possibility. The primary goal here is to replace chronically underloaded generator sets with alternators that can exceed the base load and use batteries to can cover the spikes. And now I have said too much.

 

[waross]

Transformer ratings:
Transformers are rated in Volt-Amps, (or KVA)
The current is limited by heating in the windings.
The voltage is limited by magnetic saturation, however AC saturation is frequency dependant.
The frequency/saturation relationship is described by the Volts-per Hertz ratio.
A transformer rated at 240 Volt 60 Hz has a V/Hz ratio of 240V/60 Hz or 40 Volts per Hertz.
At 120 Hz, this transformer winding will accept 480 Volts without saturating.
With the voltage doubled, the KVA capacity will be doubled.

The wide speed range is a challenge.
At low speeds, the maximum voltage is low and the frequency is low.
If the alternator is belted up enough to get a higher voltage and frequency at low prime mover speeds, it may fly apart at high prime mover speeds.

A possible solution:
Drive the alternator through a jack shaft and belt it up to give high voltage and frequency at low prime mover speeds.
Use an electric clutch to engage the jackshaft. (Similar to the clutch on an automotive A/C compressor.)
Don't engage the clutch until the speed/frequency is high enough to avoid transformer saturation.
Use a second electric clutch to drive the jackshaft at a lower ratio/speed.
When the prime mover reaches mid-range speed, around 900 RPM to 1200 RPM, disengage the high ratio clutch and engage the lower ratio clutch.
That should address the twin issues of lower output at lower prime mover speeds and possible mechanical damage at higher prim mover speeds.
And you get both KISS and KICH, Keep It Simple and Keep It cheap.
Looking at pictures of automotive alternators, I estimate 60 Hz output at 1800 RPM alternator speed.
Maximum speed is around 8000 RPM to 10,000 RPM depending on the model.
I would start my design at 7200 RPM max. or 240 Hz. Still well below the 400 Hz aircraft standard.

 

[edison123]

The standard cold rolled steel are designed for 60 Hz for the real world. For higher frequencies like aircraft 400 Hz, special thin amorphous core is needed due to hysteresis loss and eddy current loss. Those cores are mighty expensive.

 

[waross]

In order to keep a design KISS and KICH, may we accept greater losses and oversize accordingly.
Am I correct in estimating that the increase in hysteresis loss and eddy current loss will be roughly linear with frequency or will eddy current loss be a square function?

 

[LionelHutz]

A custom built alternator is likely your best and simplest bet. One that can physically withstand your approximately 3:1 speed range and also output the voltage you want over the same speed range.

Charging a battery bank and then running everything off that bank is likely much simpler than trying to combine two variable AC 480V sources.

 

[waross]

A custom built alternator is likely your best and simplest bet. One that can physically withstand your approximately 3:1 speed range and also output the voltage you want over the same speed range.

Automotive alternators will operate over a speed rang of about 10:1 to 14:1.
Automotive alternators are capable of outputting 100 Volts or more, but that does reduce the speed range.

 

[crshears]

Once heard of an aeronautic 400 Hz main engine supplied system that used a sort of torque converter equipped with governor to drive a constant speed alternator; any way to kludge something like this?

 

[TugboatEng]

There was a pair of boats built in the 1970's that had graced our fleet that had hydraulically driven generators that operated at constant speed using variable displacement pumps/motors with a governor. These supplied 35-50kW of power at 480V 3P. The system performed terribly. It wasn't responsive enough to manage the 15-20 HP load from the steering hydraulic power unit.

 

[waross]

There was a pair of boats built in the 1970's that has graced our fleet that had hydraulically driven generators that operated at constant speed using variable displacement pumps/motors with a governor

Thanks for sharing Tug.
I have been aware of those units but this is the first "Hands On" information to come to me.
Your first hand appraisal is much appreciated.

 

[TugboatEng]

Built by Dakota Creek, designed by Spaulding. Fantastic boats otherwise. I never got to deal with the systems myself as they had all been removed from the boats. The engine speed ranges were 400-900 rpm. Not much rotational energy stores to absorb load swings. I'm curious about which voltage regulators were used at the time because most of the analog units of the period had substantial under frequency roll off rates. UFRO may not have been so necessary considering a 1500hp engine was driving a 50kW generator.

 

[waross]

I doubt that they would have had UFRO.
While I had done some generator work prior to the 90's, I didn't encounter UFRO until the 90's.
Prior to UFRO it was a known issue of voltage regulators burning up if the generator was run under frequency.
The early sets always had a governor switch and the governor was never turned on until the set was up to speed.
Regulators often burned out if a mechanic with lots of diesel experience but little or no generator experience worked on the engine.
It was common for mechanics to run the engine at an idle while checking and adjusting the valve lash.
If they did that with the AVR turned on, an AVR burnout was guaranteed.