Home / Renewable Grid Power – Index / Renewable Grid Power – Baseline

Renewable Grid Power – Baseline

Case Study on Renewable Grid-Power
Electricity

Previous – Content – Next

4. The Electrical Sector Baseline

4.1 Introduction

This section presents the electrical sector baseline used
in the model and case study. To ensure consistency with
other case studies being completed by the NRTEE and
to ensure the results are comparable with other modelling
efforts (such as the AMG), the Canada Energy Outlook
– Update, 1999 (CEOU99) energy baseline is adopted.22
The discussion is organized to match the parameters
that are used in the RFF model, namely:

• Geographical coverage
• Time period
• Discount rate and dollar year
• Marginal cost of fossil fuel generation
• Baseline emissions intensity of fossil fuel electricity generation
  
• Carbon abatement cost curve for fossil fuel electricity generation
  
• Price elasticity of demand for electricity
• Required return on R&D investment
• Current expenditures on renewables R&D in Canada
• Baseline demand for electricity (kWh) from fossil fuel in two periods
• Baseline of annual renewable energy supplied (kWh) in two periods
  
• Carbon price.

4.2 Geographical Coverage

The analysis is conducted from the perspective of Canada.
While it may be analytically preferable to assess the
management instruments for each region or province and
then roll up the number to an aggregated national total,
it was agreed during the initial project meeting that
budget constraints necessarily limited the analysis
to the national level.23

Although
the case study uses aggregated national parameters for
the analysis, it is important to note that all national
data were developed from disaggregated regional and
sectoral values. Consequently, in developing the aggregated
national parameters, weighting schemes were used to
ensure that they reflect the overall circumstances that
prevail in each region.

4.3
Time Period

Based
on discussions with the NRTEE, the starting year for
applying the EFR instruments is 2010. A second period
is also modelled in 2030 to capture the longer-run impacts
of the EFR instruments and to account for R&D effects
and learning effects on renewable costs and supply.

4.4
Discount Rate and Dollar Year

To
remain consistent with analysis produced by the National
Climate Change Process (and Treasury Board), a discount
rate of 10% is used to express all results in 2000 Canadian
dollars.

4.5
Marginal Cost of Fossil Fuel Generation

Consistent
with the agreed overall case study approach, the CEOU99
was used to estimate the marginal cost of fossil fuel
generation. In this case, the CEOU99 forecast price
of electricity in 2010 was used as the marginal cost
of electricity in 2010.

As
the RFF model requires a single electricity price, the
electricity price estimates for 2010 and 2030 were weighted
to account for end-user shares of electricity demand.
This was done by using the price of electricity and
the share of electricity by end users (residential,
industrial and commercial) to develop a weighted electricity
price for 2010. The resulting weighted electricity price
was $25.70/GJ or $0.092/kWh.

4.6
Baseline Emissions Intensity of Fossil Fuel Generation

The
baseline fossil fuel emission intensity estimates the
emission reductions associated with three possible situations
stimulated by the EFR instruments:

• Reduction
  in the fossil fuel intensity
• Increased
  renewables uptake that displaces fossil fuels
• Decreased
  end-use demand due to price change in electricity.
  

In
each of the above situations, the challenge is to determine
what type of electricity is backed out or displaced
by the RET grid power. This is a very complex area that
has been the source of extensive analysis and debate
in recent years. Fortunately, this case study was able
to rely on the results of a recent (August 2003) study
conducted for Environment Canada’s Pilot Emissions
Removals, Reductions and Learnings (PERRL) initiative.
PERRL is confronted by exactly the same emissions displacement
issues as in this case study, namely: when renewable
electricity projects displace conventional electricity
sources, what type of generation (and associated emissions)
are displaced?

The
Environment Canada study was conducted using ICF’s
proprietary modelling tool, IPM (Integrated Planning
Model),24 to project the type of capacity and related emissions
likely to be displaced with the integration of renewable
energy products. The IPM distinguishes between two categories
of electricity generation: “baseload” and
“on-margin.” The PERRL analysis assumes
that renewable sources will primarily displace the fuels
used for on-margin generation. Excerpts from the PERRL
report, which identify the major assumptions related
to on-margin generation in each province, are provided
below.25

• Alberta,
  Saskatchewan, Nova Scotia and New Brunswick
  are predominantly powered by coal on the margin with
  some instances of natural gas. The situation is slightly
  different in New Brunswick, where Orimulsion plays
  a large part in total generation and is dispatched
  regularly on the margin. Historically, Alberta is
  a winter-peaking province with a coal-based grid.
  Therefore, less gas would be used during the summer,
  low-demand months as shown by the coal capacity on-margin
  in those months. In Saskatchewan, ICF assumes that
  in shoulder months coal units are less available due
  to outages for maintenance, etc. This lower availability,
  combined with import and gas prices, creates the situation
  where U.S. imports are on-margin during shoulder months.26  
• Ontario is a more diversified province in its electricity
  generation, and a mix of fuels is seen on the margin.
  Ontario is typically nuclear and hydro baseloaded,
  with mainly coal and imports on the margin. Unlike
  current conditions in Ontario, the added nuclear power
  post-2004 reduces reliance on coal and U.S. imports.
  Therefore, coal and imports are often the on-margin
  units, not baseload. The changes in Ontario’s
  on-margin capacity type – that is, bituminous,
  lignite and U.S. imports – reflect the forecast
  changes in demand (increasing over the years), environmental
  constraints (NOx and SOx tighter caps, and pending
  carbon constraints) and fuel and power prices. The
  combination of these factors shifts the on-margin
  capacity type over time.  
• British
  Columbia, Manitoba and Quebec are supplied
  power mainly through their baseloaded hydro operations.
  In British Columbia, biomass rather than gas is dispatched
  on-margin as a result of relatively low provincial
  demand and little fossil capacity. Natural gas does,
  however, cover the highest-demand periods in the winter
  months. Manitoba’s marginal power generation
  is dominated by coal, except in the peak months of
  December and January when demand becomes high enough
  to acquire U.S. imports. Quebec has very low fossil
  fuel use and therefore has very low emission factors.
  The province relies mainly on biomass and landfill
  gas for marginal generation in the high- and low-demand
  months, respectively; however, during higher-demand
  periods, U.S. imports are on-margin.

Manitoba
and Quebec are similar in their heavily baseloaded hydro
systems; however, Manitoba’s on-margin unit is
almost always coal, with the exception of one month
of importing. Manitoba is a winter-peaking province.
Therefore, December is the highest-demand month and
it is reasonable that they may have to import during
this peak period. This gives Manitoba an emission factor
of 1.02. Quebec’s on-margin units are most often
biomass and to a lesser extent landfill gas.

Quebec
does have approximately 1.5 GW of oil and combustion
turbine capacity, but they are rarely used. A Statistics
Canada publication, Electric Power Generation, Transmission
and Distribution, 1999, indicates that they have a less
than 3% capacity factor. This translates to approximately
250 hours per year. It is expected that the most conservative
influence that gas would have on the emission intensities
would be a gas value of less than 0.40 kg/kWh in the
peak month of December.

The
Environment Canada study identified the monthly on-margin
type of generation that would be displaced by increased
renewables uptake. The results were developed for each
province on a monthly basis for the 2004 to 2007 period;
the study then developed emission intensity coefficients
that corresponded with each of these situations.

As
the RFF model works on the basis of a national emissions
intensity factor, the detailed monthly and provincial
on-margin generation and emission intensities in the
Environment Canada study were used to develop a weighted
national emission intensity coefficient. The results
are shown in Exhibit 4.1.

Weighted Emission Intensity for Canada

Since
the base year is 2010 for this case study, and the Environment
Canada report only provides data for 2004 to 2007, there
were three choices for estimating the emission intensity
in 2010:

• Use
  the average of 2004 to 2007, which is 0.53 tonnes
  CO2/MWh.
• Use
  the 2007 estimate as the 2010 value, which is 0.53
  tonnes CO2/MWh.
• Forecast
  an estimate for 2010 based on a trend in the 2004
  to 2007 data, which is 0.66 tonnes CO2/MWh.

To
be conservative, this case study selected 0.53 tonnes
CO2/MWh.

4.7
Carbon Abatement Cost Curve for Fossil Fuel Generation

This
parameter is used to identify the amount of carbon abatement
for a given carbon price that will come from the electric
power sector. It is a benchmark, where firms in the
electrical sector that are faced with an emission constraint
compare the internal cost of emission reductions with
the cost of reductions from the renewables sector.

The
carbon abatement curve for the electrical sector is
based on existing plant displacement, where
emission reduction constraints force GHG-intensive generation
to be substituted with less GHG intensive generation.
Abatement costs are then estimated by comparing the
full costs of the new generation (i.e., capital and
variable) versus the variable costs for the existing
plant.

When
estimating the cost of carbon reductions in Canada,
the marginal cost of carbon reductions for the electrical
sector includes the full cost of a combined-cycle gas
plant (the typical gas plant in future periods) minus
the variable cost of coal (i.e., current coal fuel costs
must be accounted for in the carbon reduction cost).
Similarly, the incremental emission reduction from a
gas plant is simply the emission intensity for coal
minus the emission intensity for gas. The marginal abatement
cost curve is then a combination of displaced emissions
and incremental costs, where emission reductions are
supplied up to the maximum reductions available from
coal (i.e., the point at which coal is totally displaced
by gas). In reality this point would not be reached,
and indeed in this analysis the emission reduction target
is not expected to approach the total displacement of
coal.

It
is perhaps surprising that a reliable estimate of the
carbon abatement curve is not available for Canada.
That said, the incremental emission reductions and costs
that comprise the elements of the carbon abatement cost
curve can be readily estimated. First, incremental emission
reductions are estimated by province when coal plants
are displaced by gas plants under a binding EFR instrument.
The provincial differences in the incremental emission
reductions when coal is displaced for gas are provided
in Exhibit 4.2. The cumulative (or total potential)
carbon removed in tonnes per CO2/MWh is also provided.

Incremental and Cumulative Emission Reduction

Source: ICF, 2003.

Second,
Exhibit 4.3 provides the data and calculations used
to estimate the incremental and cumulative cost of CO2
reduced.

Incremental and Cumulative Costs27

* Numbers may not add due to rounding.

Source:
ICF, 2003.

The
carbon abatement cost curve is estimated by regressing
cumulative tonnes removed in Exhibit 4.2 against cumulative
cost in Exhibit 4.3. The values in the tables are expressed
as average costs but are easily converted to total costs:

From
our data, we estimate by
ordering the abatement options by region according to
cost effectiveness, and then regressing cumulative abatement
costs on .
The resulting coefficient has a very good fit and needs
only be scaled by total production to produce the parameter

to reflect the increase in marginal generation costs
from a unit decrease in emissions intensity.

4.8
Price Elasticity of Demand for Electricity

Elasticities
are useful to determine the change in end-use electricity
demand associated with electricity price increases.
In the model, the elasticity is required when the EFR
instruments affect end use by increasing the price of
electricity. If the electricity price increases, it
can be expected that there will be some demand response
to the increased price. The model is designed to capture
emission changes that stem from the changes in end use
that are attributable to the EFR instrument. Thus, a
nationally usable price elasticity of demand is required
that can be applied across the end users.

The
aggregate elasticity estimate must reflect regional
as well as end-use differences; however, such elasticities
are not readily available in Canada. Short-run elasticities
for end users found in the literature range between
-0.03 and -0.70. Based on both the literature review
and our own analysis, the short-run price elasticity
for electricity is expected to be low, indicating that
end users are not that responsive to price changes.
Over the longer term, however, price responsiveness
is much greater. Based on these considerations, a price
demand elasticity for electricity of -0.3 was selected.
This implies that a 10% price increase in electricity
will result in a 3% drop in demand.

4.9
Return on R&D Investment (ROI)

The
ROI is used in the model to capture investments by the
renewables firms in renewable technologies. Firms will
continue to invest in R&D as long as the ROI objective
is satisfied. ROI objectives of firms vary, with no
clear guidance on the level firms require. A range of
possible ROI estimates, including 30%, 40% and 50%,
was considered. To be conservative, and to reflect the
comments made during the Scoping Session, 30% was used
in the model.

4.10
Current Expenditures on Renewables R&D in Canada

This
parameter is used to account for increased production
levels (renewables uptake) that may occur when R&D
expenditures are increased. For the model, current and
projected R&D expenditures on renewable technologies
are required to forecast the value of R&D expenditures
by the renewables sector, including government expenditures,
in 2010.

According
to the latest industrial R&D survey by Statistics
Canada, R&D expenditures on renewable resources
energy technologies were $91 million in 2001.28 This includes $66 million in company self-funding, $11
million in government funding and $13 million from “other
sources.” Growth rates and projections for the
R&D expenditures of the renewable resources energy
sector are not readily available, but growth can be
inferred from a number of indicators, including company
self-funded R&D, government renewables funding,
and other sources. Further discussion of each indicator
is provided below:

• Company
  self-funded R&D expenditures in the light
  manufacturing and electric utility sectors ranged
  between 0.6% and 1% of revenues between 1999 and 2001.
  This range can be used in conjunction with a projection
  of revenue for renewable energy from the CEOU99 to
  estimate a proxy for RET R&D in 2010. Revenue
  to RETs from the CEOU99 was estimated by taking the
  weighted price of electricity (weighted for industrial,
  commercial and residential prices by their share of
  end-use sales) and the share of renewables in the
  base-case CEOU forecast. By assuming a ratio of R&D
  expenditures to renewables revenue of 0.8%, the base
  level of private sector renewables R&D investment
  was estimated to be $61 million in 2000 and $84 million
  in 2010. This implies a growth rate in expenditures
  in the order of 3.6% per year in RET R&D between
  2000 and 2010.  
• Government
  funding on renewable energy technologies
  increased 15% between 1990 and 1999, at an annual
  growth rate of 1.5%.29
  However, this number doubled between 1999 and 2002.30 When this is considered, the annual growth rate between
  1990 and 2002 was in the order of 6%. This estimate
  is used as a proxy for future growth in the government
  portion of the 2001 industrial R&D survey estimate
  (of $11 million), indicating government expenditures
  in renewable R&D to be in the order of $20 million
  in 2010. This is likely a conservative estimate given
  recent R&D funding programs under the National
  Climate Change Action Plan.  
• The
  “other sources” portion
  of renewable resources technology R&D in 2001
  is grown at the company self-funded rate of 4% per
  year between 2001 and 2010. The $13 million in other
  sources R&D would then grow to $14.9 million in
  2010.

Based
on the above indicators, total R&D expenditures
in 2010 for RETs were forecast to be in the range of
$129 million. This implies a growth of 42% in overall
RET R&D spending between 2001 and 2010, or 4% per
year.

4.11
Baseline Demand for Electricity from Fossil Fuel

The
demand for electricity generated by fossil fuels is
the reference point from which changes in emissions,
demand and renewables share of generation are estimated.
In addition, the quantity of electricity generated from
fossil fuels is a key determinant that is used to estimate
the costs attributable to the EFR instruments.

The
forecast quantity of electrical generation from fossil
fuel is taken directly from the CEOU99 and is 197,728
GWh in 2010.31
The CEOU99 does not forecast a quantity for 2030, and
therefore the case study prepared an estimate of electrical
generation from fossil fuel for 2030. Using the fossil
fuel generation data in the CEOU99, a value for 2030
was estimated using a linear regression (see Exhibit
4.4). Using the equation shown in Exhibit 4.4, which
was estimated from the share of fossil fuel generation
from 1990 to 2020, forecast fossil fuel generation was
estimated to be 316,628 GWh in 2030.

Estimated Baseline Fossil Fuel Electricity Generation
in 2030

Exhibit
4.5 provides a summary of the electricity output in
Canada that is covered under this case study. This study
covers 37% of electrical generation in Canada in 2010.

Projected Share of Grid-Power RETs and Fossil Fuel Generation
in 2010

*
Canada’s Energy Future: Scenarios for Supply
and Demand to 2025 (Techno-Vert Scenario), National
Energy Board, 2003 http://www.neb-one.gc.ca/energy/SupplyDemand/2003/index_e.htm.

**
Canada’s Emissions Outlook: An Update,
Natural Resources Canada, 1999 http://www.nrcan.gc.ca/es/ceo/update.htm.

4.12
Baseline of Annual Renewable Energy Supplied

As
with fossil fuel generation, the baseline quantity of
generation from renewables is an important reference
point for the model, as it is from this starting point
that changes in the share of renewables in the generation
mix are measured.

Unfortunately,
the CEOU99 is flawed with respect to the forecast of
renewables and an alternative source was required. Consequently,
we used the growth rate in grid-power RETs that is used
by the National Energy Board, 200332
and applied this value to the CEOU99. In the NEB, 2003,
renewables grow by 512% between 2000 and 2025. This
implies that renewables growth will be in the order
of 600% between 2000 and 2030. Applying this estimate
to the actual 1997 value reported in the CEOU99 produced
a baseline grid-power RET generation of 31,000 GWh in
2010 and 97,000 GWh in 2030, which were assumed in the
model.

Estimated Baseline Grid-RET Electricity Generation in
2030

4.13
Carbon Price

The
carbon price is set in the model to establish an emission
reduction (or environmental) target. The carbon price
is linked to the carbon abatement cost curve so that
an emission reduction target is identified at a given
carbon price. Setting the carbon price is important
for the analysis since it identifies the emission reduction
target that all of the EFR instruments must achieve.
That is, the model ensures that the instruments are
compared on a consistent basis through the identification
of a carbon reduction target. The carbon price is expressed
in dollars per tonne reduced and is set at $10/tonne.

The following section presents the case study methods
and results.