The fundamental problem with the eff2 and eff3 module is that a "a watt saved by an eff2 module, is much more expensive than a watt generated by solar/accu". Both have a upfront cost and then zero cost per watt so the investment can't ever pay off compared with solar/accu.
Eff2 and Eff3 can never be useful unless they are allowed to operate on things which consume like 1-4MW... then they could save enough watts to justify the high investment.
Turns out there is such a thing, assemblers boosted by prod and speed modules.
A potential solution
Initially my thought was that eff modules could be balanced around reducing the final or absolute energy consumption post all (other) module effects, rather than the base energy consumption, though that is iffy especially in terms of the balance between eff1 and the higher level modules.
Perhaps a better thought, is that there could be a separate parameter: "Absolute Consumption", the absolute consumption parameter would affect the final energy consumption instead of the base energy consumption.
So the modules then might look like:
- Eff1: -30% Base Consumption
- Eff2: -30% Base Consumption, -10% Absolute Consumption
- Eff3: -30% Base Consumption, -20% Absolute Consumption
In a Prod3+Speed3 setup we're looking at energy consumption of about 2MW, so directly in an assembler 3:
- eff2: -63kW + -194kW = -256kW
- eff3: -63kW + -387kW = -450kW
- eff2 in a beacon (8 assemblers): -256kW / 2 * 8 = -1MW
- eff3 in a beacon (8 assemblers): -450 / 2 * 8 = -1.8MW
Other uses for Absolute Consumption
Perhaps it could also be useful for Productivity modules, to make the energy cost more punishing when used in combination with speed beacons. For example the 3 levels could simply have +10%,+20%+30% Absolute Consumption and no Base Consumption. The lower speed already provides a pretty severe energy penalty, but that speed/energy penalty is easily mitigated with speed modules. The absolute consumption would be unmitigatable except directly by eff2/3 modules.
