Some hard numbers regading the electricity nerf

[cranky corvid]

I ended up building this test rig today. It rotates a large electric generator at 89 RPM with a 40.5:1 gear ratio using six large diesel engines. This has a stable output of 836.63 units of electricity according to the readout from the generator.

I set up a large electric motor that draws from the test rig's main electric network and powers a second large electric generator that is not hooked up to the rest of the power network. Since this electric motor draws power from the main network, when the main engine is running at maximum gear, increasing the electric motor's throttle to the value that achieves the highest stable output reading from the attached generator establishes how much power makes it through the diesel engine -> generator -> electric motor conversion, which determines how efficient electric propulsion is in the game.

Before the recent changes to the electricity system, efficiency could actually increase during the conversion process, which violated the first law of thermodynamics and enabled the creation of perpetual motion machines. Realism-wise, some loss of efficiency during conversion is of course expected.

So what's the actual power conversion efficiency with the latest hotfix? The main engine outputs 836.63 units like I mentioned above, and the maximum stable output I could manage from the large electric motor was... Roughly 250 units. In other words, running mechanical power into a generator and then converting it back into mechanical form involves an efficiency loss of roughly at least 70% compared to just running the original power source to the destination directly (in the case of large generators & motors).

So what does this mean for the game? Well, for one thing, it renders diesel-electric propulsion pointless on anything non-submersible - if you're going to have a diesel on your vehicle, you might as well connect it directly to the propulsion instead and take the tripled propulsion power output. And since electric propulsion is now extremely inefficient, any vehicle that relies on it for main propulsion is going to perform strictly worse compared to using alternative power sources, aside from cases where they are ruled out due to strict design constraints (mainly submersibles and micro-vehicles too small to carry a diesel engine, which are of course heavily nerfed by this change).

Anybody that thinks they can achieve a higher power conversion efficiency is welcome to try - if there's some magic trick that renders electric propulsion actually worthwhile after the nerf, I would love to know about it. Feel free to tinker with my test rig. Here's a list of things that did NOT substantially alter efficiency:

* A single large diesel engine produced roughly 1/6th of the conversion output compared to using all the 6 engines on the test rig.
* The large electric motor required some gearing to achieve the current peak stable output, but further gains from increasing the gear ratio between it and its generator seemed marginal at best.
* Connecting the main generator and the large electric motor to more batteries did not affect the result.
* The large generator has a maximum RPM of 100. It will not offer further mechanical resistance or electric output once the amount of power input necessary to achieve that RPM is met, and any excess power will go purely into increasing the RPM of the rest of the powertrain.

EDIT: Grammar.

[Woodspeople]

SpannerMonkey checked out the values in the files, should help.

Pre update values

Generator large output electric_magnitude "5"
Generator medium output electric_magnitude "0.75"
Generator small output electric_magnitude "0.025"

Motor large electric_magnitude "4"
Motor medium electric_magnitude "0.6"
Motor small electric_magnitude "0.02"

Post update values

Generator large output electric_magnitude "0.25"
Generator medium output electric_magnitude "0.2"
Generator small output electric_magnitude "0.15"

Motor large electric_magnitude "0.8"
Motor medium electric_magnitude "0.9"
Motor small electric_magnitude "1.0"

[SpannerMonkey]

Hi, despite the recent influx of patches those post update numbers are still the ones in game, the tweaks that have been carried out since the Electricity update must be being made within the controlling code. The only definition changes of note since then have been a massive increase to battery capacity, which isn't going to help much when in certain applications it's almost impossible to charge them properly.

[Pete]

Good Stuff.

Shall we go over some of the things which are wrong with what happens in your test rig?

First off, the efficiency is roughly the inverse of what it should be. A normal mechanical-electrical-mechanical cycle is about 70% efficient, or to put it another way, has 30% losses.
This is worse than a typical mechanical system by some way, which is 85% for a car gearbox for example. This is why real life uses of electrical transmissions are restricted to situations where the layout, the sheer torque involved, or other difficulties demand its use.

Secondly- In real life, a generator has a small amount of mechanical drag from its bearings. If no current is being drawn from the generator, this is the only torque required to turn it. When current is being drawn from the generator, the torque to turn the generator is an equal amount of energy to that current, plus the losses in bearing friction, plus all the other losses in the conversion of mechanical motion into electricity.


So, what would we expect to see happen in your test rig?
Well, since the output generator is connected to nothing and is freewheeling with no current draw on it, its mechanical resistance is very low.
This means the motor driving it requires very little current to turn it at any given speed.
Which means the generator powering that motor has very little current drawn from it, and thus presents the diesel engines with very little mechanical drag.

In real life, if we connected that second output generator to a half flat battery, or an electric heater, or complete the circuit with anything of comparitively low resistance, the current draw on the generator would shoot up, increasing the mechanical drag on motor, causing the motors current consumption to increase, causing the primary generator to increase drag on the diesel engines, causing the diesel engine to burn more fuel.

This is a simplified analysis, with a few ballpark figures and things glossed over, but I'm not trying to write a textbook here. Perhaps a GCSE physics textbook level of explaination. However, if the game was true to life to that degree, it would be quite adequate really.

[cranky corvid]

Now that I'm less sleepy, one additional point. For the purposes of the efficiency measurement, the power from the main generator is the input. It then goes through an electric motor and another generator which produces the test output. This makes the test an accurate measure of the efficiency loss in an electric drive system - there's no doubling up of generators for the purposes of the measurement since the first one is the input.

[cranky corvid]

I updated the test rig to account for the latest balance changes. The conversion efficiency of the test rig now tops out at a much more reasonable 76% rather than 30%.

The output of the input generator more than tripled because of the efficiency boost, but the 6 large diesel engines now run the generator at only about 50 RPS compared to the slightly over 90 from before, so even beside the efficiency change, it looks like the throughput ceiling of generators has greatly increased.

This does mean that pre-nerf designs will still need changes even though the balance is much more reasonable now, since the amount of generators needed relative to engine power is far lower compared to pre-nerf - since generators are more efficient at higher RPS, the excess generators in pre-nerf designs drag their efficiency down.