[LDES-coremodel] Update of existing and planned generation for SWITCH WECC db

[Julia Szinai]

Hi all,
I'm almost done updating the script and pushing the data of the existing
and planned generation in the US portion of the WECC into the SWITCH
database. The updated scripts 1) scrape the data from EIA forms 860 and
923, 2) process and standardize the data, and 3) upload the data into the
database. They build on the original scripts in github here:
https://github.com/RAEL-Berkeley/eia_scrape. (I haven't pushed my updated
code to github yet).

This data contains the generator parameters that go in the main
generation_projects input file into SWITCH (including heat rate, capacity,
build year, max age, etc), as well as the variable capacity factors for the
renewable generators. The data is as of 2018 (since that is the most recent
year for which there is both generator capacity and annual energy data),
but I've removed generators which have been retired between 2018 - May 2020
(since retirement data is more current). I also added planned retirement
dates whenever available.

I compared the total capacity in the WECC between the previous data of
existing generation (as of 2015) with this update, and with the total
current capacity in the WECC according to the WECC website (
https://www.wecc.org/epubs/StateOfTheInterconnection/Pages/Capacity.aspx).
The updated data has more solar, wind and gas than the 2015 data
(interestingly also more coal), and the numbers are pretty close to that of
the WECC website. I've attached an Excel file of the comparison.

I have a couple of questions for the group that I think need to be settled
before I finish preparing the data for a baseline scenario:
1. The variable capacity factors in the original script are added to the
"variable_capacity_factors" table in the database. However, the inputs for
the WECC SWITCH runs use the capacity factors from the
"variable_capacity_factors_historical." What is the relationship between
these and is there an additional script I'm missing? More importantly, it
appears that the variable capacity factors for solar are off by 7 hours
(likely related to time zone). The adjustment in the original script
doesn't look like it is properly correcting for this, and I think it is
still mismatched in the final data going into SWITCH. This could be the
reason for the infeasibility of the 100% renewables run, if for example,
solar is not generating during the day because the hours are incorrect. Is
there some SWITCH adjustment for timezones, or time points that I'm missing?

2. hydro capacity factors: I've added monthly hydropower data to the
database for 2004 - 2018. For the baseline scenario, should I calculate
average monthly capacity factors across those years? This would reflect
"average" operations under dry and wet years. The previous data only
included hydropower for 2010 - 2015, so not as representative of the range
of hydroclimatic conditions.

3. Canadian + Mexican parts of WECC: The data I scraped is from the EIA, so
it doesn't have data for existing generation in Canada (BC and Alberta) and
Mexico (small part of Baja CA). Looking at the prior data in the database,
the generators for these load zones were all "proposed" and there was no
existing generation. I guess SWITCH just "built" all the
existing generation in the first period to meet the load? I'm not sure but
that seems a bit weird, and could cause overestimated costs for the first
period. Am I missing something? Should I look for updated existing
generation data for BC and Alberta, and part of Mexico? The query I ran
below came out empty (8 and 9 are the Canadian load zones, and
generation_plant_scenario_id 14 is the scenario used in the last CEC
report):

SELECT *
FROM generation_plant
JOIN generation_plant_scenario_member
USING (generation_plant_id)
WHERE generation_plant_scenario_id = 14
And name != 'Proposed'
AND load_zone_id in (8, 9)

4. This data update was just of the existing generation plants. The
baseline scenario will need to also include the data from candidate
generators. My plan was to append the list of candidate generators from one
of the scenarios used in the prior CEC study. However, there are several
scenarios of candidate generators. As we discussed some of the prior
scenarios have candidate renewable generators excluded because of
environmental restrictions, based on data from E3 I believe. Do you know if
we will get an updated set of environmentally-friendly candidate generators
from E3? If not, should I use the restricted scenario of candidate
generators for the baseline?

5. The existing battery projects that are in the generation_plant_cost
table for the new scenarios 19 and 20 will need to be updated in the
storage_energy_capacity_cost_per_mwh to reflect whatever is the baseline
assumption for battery costs. Right now I've left these as NULL in the
table.

6. The capacity factors for solar were previously averaged by load zone
across residential, commercial, and utility-scale solar PV. I did the same
for this data, but not sure if we want to revisit this simplification or
keep the residential and commercial vs. utility-scale solar capacity
factors separate.

*Julia Szinai*
PhD Candidate | Energy & Resources Group | University of California,
Berkeley
Graduate Student Researcher | Lawrence Berkeley National Lab
NSF InFEWS Fellow
Energy & Resources Group, MS '17
Goldman School of Public Policy, MPP '17
University of California, Berkeley
jszinai at berkeley.edu
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ucmerced.edu/pipermail/ldes-coremodel/attachments/20200917/e2d73151/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Compare capacity by energy source for new and old gen plant scenarios to WECC website.xlsx
Type: application/vnd.openxmlformats-officedocument.spreadsheetml.sheet
Size: 14241 bytes
Desc: not available
URL: <http://lists.ucmerced.edu/pipermail/ldes-coremodel/attachments/20200917/e2d73151/attachment-0001.xlsx>