12. Appendix


Effective capacity is a measure of a given generator’s capacity that the Alberta Electric System Operator can rely upon to be available during the peak load hours, such as early evenings in winter. The operator allocates different effective capacity factors to different generation sources — for example, most firm sources are assigned 100 per cent, while wind is assigned 20 per cent (AESO, 2012).

Using the peak-load projections under our Continued Fossil Reliance scenario, we refined the Clean Power Transition and Clean Power Transformation scenarios until we attained at least a 15 per cent reserve margin in each year. With considerable wind and solar additions—especially in the Clean Energy Transformation scenario—this spurred expectations for a robust energy-storage sector in the province but also sometimes required more quick-ramping capacity in the form of simple cycle natural gas capacity.


We engaged Solas Energy Consulting Inc.—a project and business development firm with offices in Calgary and Colorado—to model the impacts of each of our three scenarios on wholesale energy prices. We asked the firm to do so for each of the years 2023 and 2033.

Under the current market design, wind, solar and geothermal power generation do not offer into the merit order. Rather they are included as a negative load and reduce the total demand dispatched through the merit order. Pembina Institute provided the generation capacities for each technology type.

Wind power generation was forecast by adding capacity to diverse geographic areas based on Solas’ experience in the wind industry. A time series of wind power generation was created for each location where wind arms are likely to be located. The individual regional wind power generation time series were then combined to create a total wind generation time series.

Solar power generation for the Clean Power Transition and Clean Power Transformation scenarios was calculated from a time series provided by the NREL solar resource tool and multiplying by installed capacity.

Geothermal power generation capacity was provided by the Pembina Institute and was assumed to operate as baseload at a 70 per cent capacity factor.

We assumed energy efficiency measures would be implemented in proportion to historical demand. That is, efficiency contributions were higher during hours when demand was higher, and the contributions were lower when demand was lower. Improvements in energy efficiency reduced the demand rather than having an effect on the merit order.

Pembina provided the capacity of energy storage for each scenario. Energy storage was implemented in the merit order by inserting bids to purchase energy at low power prices and offers to sell energy at higher prices. Bids and offers for energy storage were structured such that the revenue was sufficient to recover the cost of purchased energy plus efficiency losses. Further, storage bids and offers were structured such that total injections and withdrawals balanced over the course of the year and storage did not affect the total annual energy demand.

A demand time series was then created for 2023 and 2033 by increasing the demand in each hour by the respective demand growth assumption. This was done to maintain correlation between wind generation and demand.

Power prices were determined for each of the six scenarios based on the dispatch and demand for each hour in 2023 and 2033.


We modelled the greenhouse gas results of our three scenarios using empirical or widely accepted intensity factors for each fuel source, gathered from federal greenhouse gas emission reporting data and unit-generation data. Because cogeneration produces both steam and electricity from the fuel combustion, it poses a special allocation problem for determining what proportion of the combustion emissions are allocated to electricity generation. The Independent Power Producers Society of Alberta’s recent report (EDC Associates, 2013) used a 0.25 t/MWh intensity for cogeneration’s electricity emissions. While not endorsing this allocation, we employed the same intensity for consistency in the discussion around different GHG emissions from different scenarios. Moreover, 0.25 t/MWh falls within the range of intensities derived from different allocation methods employed by Doluweera et al. (2011).

Because of the large overall emissions and large variability between units, coal unit intensities were assessed individually—for each unit—using empirical data, accounting for recent or planned incremental capacity increases to existing coal units as efficiency improvements.

We also assigned carbon intensities to the transmission interties connecting Alberta with British Columbia, Saskatchewan and Montana, based on the province or statewide carbon intensity of those jurisdictions’ respective grids.