jrissman
commented
3 years ago Seems reasonable to me. The current structure was built assuming the electricity generated any given industry would not exceed what that industry consumes (summed across all individual facilities/companies within that industry). But it sounds like there are places where that is not true, and I agree that having electricity exports from industry to the grid supported in model structure would be needed to handle this situation correctly. This will also improve cash flow accuracy. And accurate cash flows are what feed into our jobs and GDP calculations.
I'm not sure what you mean by "do it for heat also." You mean an industry produces more heat than it uses, so it sells the heat to... whom? Is this done to a significant degree? Heat is less able to be transported over long distances or stored than electricity is.
As far as other energy transformations, the two that come to mind are refining (converting crude to secondary petroleum products) and coke ovens (converting metallurgical coal to coke). I don't think the Chemicals industry is an example. They do use a lot of fossil feedstocks, but their products are not generally fuels/energy. They produce things like fertilizers, plastics, solvents, adhesives, paints, personal care products, etc.
Is there some capability you need with respect to refineries or coke ovens that we don't have today, in terms of policies you want to model or outputs you want?
robbieorvis
With additional state models and now a model for the EU, one thing we are consistently running into are challenges around how industrial CHP is currently handled in the model. In particular, there is a lot of industrial CHP being sold back to the electricity grid in industry-heavy states and in the EU. Right now, we don't have a great way to handle this. We are mostly putting the energy consumption for industrial CHP plants in the industry sector, offsetting electricity consumption by the amount produced by industrial CHP, and setting the input data values for the share of heat coming from CHP so that the heat demand from dedicated heat plants is correct. In instances where the are is net production of electricity in the industry sector (ie electricity is being produced in excess of what is needed and sold back to the grid) we are attempting to repurpose a power plant type to represent industrial CHP, which has significant limitations (e.g. limited to picking a single fuel and ensuring that the repurposed power plant type is not hard linked to an industry sector).
It would be a significant improvement in the industry and electricity sectors of the model if we could separately track and model industrial CHP and if these plants were eligible to output power to the grid. Their output might be fixed based on the capacity given the fact that they are being used for heat and producing electricity as a byproduct, kind of like using guaranteed dispatch.
If we decided to make this enhancement, it would probably make sense to do it for heat also, since the two can be tracked together. In fact, we might use the opportunity to build out a part of the model that tracks transformation of primary to secondary energy consumed on-site in the industry sector. This would primarily capture heat and electricity produced in industrial CHP plants and also the use of crude byproducts in refineries (and maybe the chemicals sector). I don't think it would significantly increase the data requirement since much of this data is already gathered and has to be very carefully removed in the input data files (which itself is prone to errors). It may actually simplify things to have the model doing the calculations instead of trying to do these outside the model in Excel.