Functionality of the RPS constraint

[xavier]

Hello, 
 
I am finding it hard to understand the functionality of the RPS constraint. In my model, results under the RPS scenario, I see that the percentage share of generated electricity with renewable sources is well above the minimum RPS percentage defined in the scenario file. To confirm if my model has an issue, I did run the Starter model, and its results showed the same trend.  Can you please explain this behaviour?
 
I assume the RPS input file defines the minimum percentage share of electricity that should be generated with renewables, and once this minimum share is achieved, the model should pick the other cheaper non-renewables. As an example, if RPS = 30% then we should have:

Total electricity generated with renewables >= 0,3 * Total electricity generated such that once the 30% renewable share is achieved, the rest of the share will come from the more competitive sources (like those prioritised under BAU) and not the expensive renewables. 
Is this correctly understood?
 
Also, running the Starter model's RPS (without the UC-90 constraint) gives total generated electricity lower than that generated in the Reference scenario.  Can you also please explain why?
 
Best Regards,
Xavier

[Antti-L]

If I understand correctly, your question is not related to VEDA-FE as such, but to a specific model, the VEDA TIMES-Starter model and the Renewable Electricity Portfolio Requirement (Scen_P-RPS) included there. Therefore, you might wish to post the question under the VEDA-TIMES Starter Forum, or you could contact the model developers, the DecisionWare Group.

I have not been using that model for other than basic testing, and never with the RPS scenario. But from a general point of view, I can try to answer the final question:

> Also, running the Starter model's RPS (without the UC-90 constraint) gives total generated electricity lower than that generated in the Reference scenario.  Can you also please explain why?

If the results are different when including the RPS scenario, the RPS constraint must be active, i.e. binding in at least some of the model years. In other words, the share of renewable electricity in total electricity generation is not meeting the RPS target shares when not including the RPS constraint, because if the targets would have been already achieved, adding the RPS constraint would have no effect. And when you add any constraints that are binding, the value of the objective function can only increase, it can never decrease. Consequently, the total system costs are then higher, and because the constraint focuses on electricity generation, the cost increase is reflected in the equilibrium prices of electricity, which are higher under the added RPS constraint. The endogenous demand curves for any commodities in the model normally have price elasticities, meaning that when the price increases, the demand for the commodity decreases. And that is the explanation for your lower total electricity generation.

[xavier]

If I understand correctly, your question is not related to VEDA-FE as such, but to a specific model, the VEDA TIMES-Starter model and the Renewable Electricity Portfolio Requirement (Scen_P-RPS) included there. Therefore, you might wish to post the question under the VEDA-TIMES Starter Forum, or you could contact the model developers, the DecisionWare Group.

I have not been using that model for other than basic testing, and never with the RPS scenario. But from a general point of view, I can try to answer the final question:

> Also, running the Starter model's RPS (without the UC-90 constraint) gives total generated electricity lower than that generated in the Reference scenario.  Can you also please explain why?

If the results are different when including the RPS scenario, the RPS constraint must be active, i.e. binding in at least some of the model years. In other words, the share of renewable electricity in total electricity generation is not meeting the RPS target shares when not including the RPS constraint, because if the targets would have been already achieved, adding the RPS constraint would have no effect. And when you add any constraints that are binding, the value of the objective function can only increase, it can never decrease. Consequently, the total system costs are then higher, and because the constraint focuses on electricity generation, the cost increase is reflected in the equilibrium prices of electricity, which are higher under the added RPS constraint. The endogenous demand curves for any commodities in the model normally have price elasticities, meaning that when the price increases, the demand for the commodity decreases. And that is the explanation for your lower total electricity generation.

Thank you Antti-L for your response. I think the model optimises generation expansion while meeting a set of constraints. One of the constraints is that the generated electricity must meet demand at any given time-period. Therefore, a lower level of total electricty generation below the demand level fails to staify this constraint, particularly where we have not activated the elasticity of demand to its prices function, demand side energy efficiency or the UC-90. However, let me transfer the question to the Starter model forum. 

Best Regards,
Xavier

[Antti-L]

Well, if the solver reported an optimal solution, the lower level of generated electricity was indeed meeting the demand at any given time-period; otherwise the solution would have been infeasible. And so also the demand for electricity must have been lower.