Question: can we take normal conversion losses into account?

[ChaelKruip]

This issue is just to clear up some confusion.

Capacities

In ETMoses, we have capacity and carrier_capacity. It is my understanding that these correspond to the ETE equivalents: output_capacity and input_capacity. The possible conversion loss of a technology that turns one carrier into another is given by

loss = 1 - efficiency

where the efficiency is defined as

efficiency = capacity/carrier_capacity

performance_coefficient

As far as I understand this, the performance_coefficient in ETMoses equals the coefficient_of_performance in ETE which is 1.0 for all technologies except heat pumps.

In ETMoses it is used to translate between capacity and carrier_capacity as in

carrier_capacity = capacity / performance_coefficient

I think this works correctly for heat pumps but results in capacity being (*wrongfully) equal to carrier_capacity in all other technologies.

What to do?

@RobTerwel has researched which numbers would be correct as capacity and carrier_capacity, taking into account conversion losses. Two notable exceptions are P2P (batteries) and EV which actually need to lose energy twice (both when charging and when discharging).

@antw what do you think would be a good (temporary) way to initialise capacity and carrier_capacity in ETMoses?

Some options that come to mind:

• Add a column to the database with carrier_capacity (or efficiency) and initialise from there
• Extend the ETE communication functions to import both 'input' and 'output' capacities

[antw]

I'd prefer to do it right, and import the input and output capacities directly from ETEngine. I'm not convinced that is sufficient to model loss, however.

In Moses, a capacity is a bottleneck which limits the (maximum) flow of energy, rather than a conversion which may have loss. For example, a hypothetical P2P battery with an input capacity of 2kW and output capacity of 1kW charges at full capacity for one hour. 2kWh is stored. It will take two hours to fully discharge (2kWh stored / 1kW output capacity = 2 hours). It does not mean that 1kWh is discharged and 1kWh is thrown away as loss.

My feeling is that we want to support input and output efficiency in addition to input and output capacity.

The carrier_capacity attribute is an ugly solution for the lack of an ambient_heat input for the heat-pumps; the only place I can guarantee it works is in the Buffer class which implements the heat-pump behaviour. I would like to be rid of it in favour of a proper capacity/efficiency system.

[ChaelKruip]

In Moses, a capacity is a bottleneck which limits the (maximum) flow of energy, rather than a conversion which may have loss.

Yes. But (!) doesn't it behave as loss for heat producing technologies for instance? If a households_space_heater_network_gas produces heat at its maximum output_capacity of 22 kW, doesn't it consume gas at 27.5 kW (22 kW/ 0.8)? The discrepancy of 5.2 kW can be interpreted as loss right?

My feeling is that we want to support input and output efficiency in addition to input and output capacity.

As a longish-term solution I fully agree. But currently, the numbers are wrong and I would like to be able to come up with a temporary solution to allow current users to get somewhat physical results :grimacing: