Ssc 599 negative cleared capacity

[brtietz]

Improve handling of cleared capacity throughout merchant plant code, and some bugfixes for the battery time series prices.

• Only use positive revenue values in revenue row of cashflow table, to match other financial models
• Negative values of revenue are now assigned to cf_utility_bill, with new consumed_cost outputs
  • Note that the previous code had a copy-paste error vs the other financial models, and was off by a factor of 1000 for a kW to MW conversion, hence the large increase in NPV in the two unit tests.
• common.cpp now plays more nicely with cleared capacity as a percent of generation, so that dispatch is the same between single owner and merchant plant.
• Users get a warning (not an error) if the negative cleared capacities don't match generation

The result is that given the same price series for one year, merchant plant and single owner now (once again?) produce the same financial results. Test file here:
aligned price series.zip

Differences in the "all markets" case are worth some debate, but I worried they'll be late for the patch, so I created another issue here: #825

[sjanzou]

Great that the SO and merchant plant results are aligning!

Two issues:

1. "capacity_factor_sales" is missing from results in Summary tab in SAM, e.g.

2. The default results have changed significantly for SAM - please review ssc_599_negative_cleared_capacity branch of SAM and pull request Updated Windows defaults results based on ssc pull request 826 SAM#1058

[cpaulgilman]

Looks good to me aside from the missing capacity_factor_sales issue Steve raised.

[brtietz]

capacity_factor_sales shows up now that #819 has been merged into patch.

@sjanzou Thanks for re-running the defaults! The prior LCOEs were overestimating the cost of grid energy by a factor of 1000 (missed kWh to MWh conversion) so the large decreases are expected.