CSP gen cost information

[Nicoc92]

Hi,

Upon reviewing the output data from the GCAM model and looking at the attached query on 'elec gen cost by tech,' I don't understand why the cost of the CSP_storage technology is significantly lower than that of the regular CSP, being more than 20 times cheaper.

Do you have any idea why ?

Thanks in advance,

[ssmithClimate]

It looks like you are examining a global query result - you will need to look at costs for a specific region. Those will make more sense.

For CSP, capital costs of CSP with storage are higher than CSP without storage, but the capacity factor of CSP with storage is higher. As a result, the overall electricity cost for CSP with storage is slightly less expensive than CSP with storage.

These will vary by region because the solar resource varies by region.

[Nicoc92]

Thanks for the quick response!

I just reviewed the structure in the XML file (electricity_water in global technology), however, I am not sure I fully understand what the backup-capital-cost, backup-capacity-factor, and capacity-limit-backup-calculator correspond to in the intermittent technology for example. Additionally, I am not entirely clear on the sources. Could you please provide further clarification?

Thanks in advance,

[pkyle]

The cost of CSP can be disaggregated into 3 components: (1) the standard non energy cost, which is the sum of levelized capital, fixed O&M, and variable O&M; (2) The cost of the “global solar resource”, currently a flat unlimited resource supply curve at a nominal price ($0.001/GJ); (3) and the backup/intermittency related costs. The third one is not written out to the database, but can be estimated as total technology cost minus the non-energy costs (“cost by tech and input”) and the nominal global supply cost. In the input XML, the unit of backup capital cost is (1975) dollars per kilowatt per year of installed backup turbine capacity; the assumption is equivalent to a capital cost times the fixed charge rate. This is described in the header comments of input/gcamdata/inst/extdata/energy/A23.global_inttech.csv and the model code for the function is in cvs/objects/sectors/source/capacity_limit_backup_calculator.cpp. As the share of renewables on the grid increases, so does the ratio of backup capacity to additional renewable installed capacity. At the limit, each additional GW of installed renewables requires an additional GW of combustion turbines. There is also an operating cost to the combustion turbines, which is calculated from an assumed capacity factor of 5%, exogenous O&M costs, and the costs of the purchased fuel. Emissions penalties may also apply. In any case, replicating the model’s estimated backup related costs in scenarios where these costs are not trivial would take some work, which is why I recommend that if you do want to know what these backup related costs are, you should derive them from the other reported costs.

[Nicoc92]

It looks like you are examining a global query result - you will need to look at costs for a specific region. Those will make more sense.

For CSP, capital costs of CSP with storage are higher than CSP without storage, but the capacity factor of CSP with storage is higher. As a result, the overall electricity cost for CSP with storage is slightly less expensive than CSP with storage.

These will vary by region because the solar resource varies by region.

In reviewing the data across different regions, I’ve found that the cost of CSP without storage is actually always higher than CSP with storage.