Alternator efficiency at higher temperatures

[Simon]

I am told that the hotter an alternator gets the less efficient it becomes. Is there any generic information out there about what it is likely to be ?

[Andy Watson]

Largely copper losses I believe. The resistance of the copper increases (about 0.4% per degree ?) leading to \$i^2 R\$ losses.

[Richard Head]

But the voltage drop across the rectifier diodes decreases (-2.2mV/deg C) as load increases so may offset the increased losses due to winding temperature.

[MagicSmoker]

I am told that the hotter an alternator gets the less efficient it becomes. Is there any generic information out there about what it is likely to be ?

You might be surprised at just how inefficient the typical claw pole wound rotor alternator is in the first place: about 55%.

Otherwise, the temperature coefficient of resistance for copper is +0.4%/C, and engine compartments are hot places... So, yeah, pretty crappy all the way around.

[Simon]

This will be a few hundred amps in 80C

Sent from my Moto G (4) using Tapatalk

[Gyro]

This will be a few hundred amps in 80C

I suspect that you're going to need to provide a bit more context and detail on this one.

[Zero999]

Largely copper losses I believe. The resistance of the copper increases (about 0.4% per degree ?) leading to \$i^2 R\$ losses.

But the voltage drop across the rectifier diodes decreases (-2.2mV/deg C) as load increases so may offset the increased losses due to winding temperature.

It looks like it can be calculated, if you know the resistance of the stator and the voltage drop across the diodes at a known current and temperature.

It would be interesting to see which effect will dominate.

[james_s]

80C doesn't sound out of line for typical operating conditions, I've never measured the temperature of the alternator in my car before but it's on an aluminum bracket just a couple inches from the engine block and only inches below the exhaust manifold and turbocharger. I would actually be surprised if it doesn't go higher than 80C and it seems to work fine. A couple hundred amps is higher than most though, 100A alternators are somewhat typical, 140A is getting fairly big.

[Simon]

There is a difference between ambient and internal temperature, I am talking about an ambient of 80C

Sent from my Moto G (4) using Tapatalk

[Simon]

Again this is not a car, many things have combustion engines in them. Cars have a huge hole in the bottom of the engine bay. not all installations of an engine are like this. I expect air will be drawn into the engine compartment as it will need to be blown through the radiator but when you have to allow for ambient conditions outside to be at up to 50C (standard temperature for operation anywhere in the world) that's only 30C of local heating before you are at 80C in the vicinity of some components.

[max_torque]

It's hugely dependent on alternator type and design.

Engine compartment ambient air temperatures can easily hit well over 100 degC after a keyoff hot soak, and radiant heat can be significant near exhaust line components etc

A lot of high power alternators (>200A) are now water cooled by the engine cooling loop, and these tend to operate at a higher average temperature because most of the time they are heated by that loop, rather than cooled!

Efficiency for a claw pole alternator (rather than a PM one) is actually pretty flat with temperature, as previously mentioned the increased copper loses in the rotor and stator are offset by reduced diode losses.  Certainly, you are talking  about changes in the 1 to 5% region or less.  And as a claw pole machine is really rather in-efficient overall (~65% at best) those small extra losses are pretty irrelevant!

The latest electronically commutated PM alternators are far far better (~95%) and exhibit almost no reduction in efficiency with temperature as they start with very low copper losses in the first place.  They however do have a critical "must not exceed" temperature to avoid permanent de-magnetisation of the PM material.

[Simon]

I take it that by PM you mean permanent magnet. Yes I was thinking that the efficiency of claw pole alternators is pretty awful. I have had at the back of the mind a "should we talk to a certain motor manufacturer" , But how would a permanent magnet alternator deal with changes in speed ? surely not a SMPS on the output ? or do they have to run at precise speeds ?

[max_torque]

yes PM as in Permanent Magnet.

Most alternators are currently separately excited claw pole designs, for low cost, easy assembly. Which is fine at lower power requirements, but once above around 180 to 200A, the thermal issues become tricky to solve at the low inefficiencies intrinsic with that design.

High power alternators are moving to PM, but they are pretty expensive, and of course require electronic commutation to operate at all times in order to control the output voltage / current. (even if you want zero output, the commutation must apply a fwd voltage to balance the back EMF). Control strategy would typically be an inner FOC for the magnetics, commanded by an outer loop targeting output voltage or current.

[Simon]

Sadly it is hard to google permanent magnet alternator without getting the result polluted with free energy bullshit

[max_torque]

Very few manufactures have a mature automotive PM alternator product.  iirc, Valeo do, as do Lear.   I've seen some marketing waffle from BOSCH and Denso but the product is i think vapourware in both cases!  (or certainly not on general release)

[MagicSmoker]

Sadly it is hard to google permanent magnet alternator without getting the result polluted with free energy bullshit

Look, instead for "belted alternator starter hybrid" and "pm synchronous generator automotive".

[Doc Daneeka]

It's hugely dependent on alternator type and design.

Engine compartment ambient air temperatures can easily hit well over 100 degC after a keyoff hot soak, and radiant heat can be significant near exhaust line components etc

A lot of high power alternators (>200A) are now water cooled by the engine cooling loop, and these tend to operate at a higher average temperature because most of the time they are heated by that loop, rather than cooled!

Efficiency for a claw pole alternator (rather than a PM one) is actually pretty flat with temperature, as previously mentioned the increased copper loses in the rotor and stator are offset by reduced diode losses.  Certainly, you are talking  about changes in the 1 to 5% region or less.  And as a claw pole machine is really rather in-efficient overall (~65% at best) those small extra losses are pretty irrelevant!

The latest electronically commutated PM alternators are far far better (~95%) and exhibit almost no reduction in efficiency with temperature as they start with very low copper losses in the first place.  They however do have a critical "must not exceed" temperature to avoid permanent de-magnetisation of the PM material.

Why is a claw-pole machine specifically any less efficient than any other configuration?

[SeanB]

Just an aside is that often the alternator output is designed to droop with temperature, so as to cloer follow the battery temperature coeficient, so the cold output voltage will be 14V4 and drop to 13V8 or lower as the engine heats up, so as to not overcharge the battery, which these days has a limited overcharge capacity due to there often being no  easy way to introduce fill up water in them, thus if the hot charge voltage is too high they boil off water as hydrogen and oxygen gas, depleting the electrolyte.

i notice this on my car, cold the battery voltage is 14V4, dropping to 13V5 when hot after a long (30km) drive, measured with a meter that is effectively across the battery terminals. Turn on the lights and the droop is to 13V0 after a while, from that extra 30A draw of the lights. Turn on the rear demister and the loom droops 0.4V, as that is my tap off point, and the 0V4 is the drop in the thick feed wire from the fuse box, however compensating for that the battery does stay at 13V0.

Best for Simon though is to either spec 2 alternators ( if space is available to put one in there, though whoever has to maintain that vehicle will forever curse you, your ancestors and give speculation as to your state of mind in doing such) and make them lower power units, arranged to feed either a common bus for the battery, or better yet 2 batteries, one solely for engine and accessory power and not for use with any engine off loading, and the other battery and charge sytem for any loads that are engine off or which are not engine related. As well I really would recommend a pair of electric cooling fans for the engine bay, one running off a thermal switch that operates at 70C, and the other at 30C, so that you have cooling air at all times. Switches probably best places high up in the bay, with air inlet near the bottom one side, and exit high up on the other, Screen both inlet and outlet against debris and rocks, using a heavy duty screen on both, so you keep stones and sand out of the fans.

[Circlotron]

Why is a claw-pole machine specifically any less efficient than any other configuration?

I think it is not so much because of the claw pole design but the method of achieving current limiting. The cross section of the stator iron where the windings wrap around is only just thick enough to support the maximum rated current and above that the iron saturates, preventing any further increase in output current. In normal use that thin iron is still worked pretty hard and so has high losses. That's what I read somewhere anyway.