IEEE Power & Energy Magazine - January/February 2021 - 67

units (e.g., diesel generators) are able to recover their investment costs. Otherwise, incentives to the entry and exit of
capacity work in the same manner in both examples. Note
that in the situation depicted in Figure 1(b), there is a more
extreme discrepancy between peak spot prices (potentially
very high scarcity prices) and off-peak prices (virtually
zero), which provides incentives for demand-side resources
to adjust consumption, achieving a similar effect to increasing or decreasing generation capacity in the long run.
Hence, conceptually, there is nothing that prevents the
application of the classic spot-pricing theory to systems
with high shares of generation from resources with zero
marginal cost. As we show in Figure 1(a) and (b), the only
difference is that, in systems with a lot of generation from
technologies with zero marginal cost, scarcity pricing
becomes the main mechanism to ensure cost recovery in
the long run because spot prices are likely to be zero for
extended periods of time. If liquid financial markets used
to hedge the price-volume risk over different time frames
are in place, the optimal capacity-expansion mix is secured
(and able to be financed). In these situations, consumers can
also define their optimal reliability needs and participation
in the market as an active demand response based on private
preferences (e.g., risk aversion).
In mature electricity markets following these design principles, spot prices can increase dramatically during scarcity times due to high price caps (as seen in Australia and
Texas). In practice, however, the administrative estimate of
the value of lost load (VoLL) used to determine price caps
is sometimes driven by political instead of technical considerations, which can introduce distortions. A relatively recent
development regarding price formation in periods of scarcity
(which is addressed in the following section) has been the
implementation of sloped operating reserve demand curves
(ORDCs) employed in some markets in the United States
and Mexico, where price-dependent curves replace vertical
demand curves for operating reserves.
In practice, low price caps, illiquid financial markets for
long-term contracts, and the lack of demand response can pose
real challenges for electricity markets in their purest form,
which choose to rely solely on spot pricing-including scarcity
pricing-to provide expansion incentives. These challenges are
especially pronounced for countries with fast load-growth rates
or increasing levels of decommissioning of existing generation
capacity, where the lack of new supply may result in shortages.
Furthermore, it is likely that these challenges will become even
more pronounced with increasing shares of generation from
renewables. Not only is there a tendency toward a feast-or-famine situation with regard to equilibrium prices, as illustrated in
Figure 1(b), but the technological disruption of renewables has
profoundly altered the landscape of expectations for the electricity sector. In particular, there are significant uncertainties
regarding the rate at which the cost of renewables will continue
to fall and their share in the expansion mix will continue to rise
as well as the rate at which additional innovations, such as the
january/february 2021	

emergence of distributed energy resources, demand response,
and storage technologies, will be disseminated. This combination of spot price volatility and uncertainty with regard to the
future evolution of the system creates an environment that may
threaten potential investments and loans, creating an even
greater motivation for long-term markets for financial and/or
reliability products.

Spot Pricing in Hydro Systems in Latin
America: One Form of Scarcity Pricing
Latin America is formed by 16 countries and has a power
system with roughly 400 GW of installed capacity, where
hydropower accounts for approximately 50% of the generation mix. Load growth rates have historically hovered
near 5% per year in a region where energy consumption is
roughly 1,500 TWh/year. Figure 2 shows a general depiction
of the main wholesale market design elements.
On the pricing side, only Colombia and the Central American regional electricity market have adopted a bid-based
scheme for generation dispatch and spot price formation, as
represented in shades of red in Figure 2. All the other countries in the region utilize cost-based arrangements, where generators report only their directly attributable marginal production costs (i.e., fuel costs) to build the merit-order curve for
the dispatch and pricing of electricity done by the ISO. Water
values are used as proxies for marginal production costs for
hydro plants, which are calculated by the ISO based on a set of
administratively defined assumptions and with the aid of stochastic optimization models. Given the cost-based merit-order
curve, spot prices are defined as the cost of the marginal unit
needed to meet demand in each settlement period.
Although cost-based markets have some disadvantages
compared to bid-based ones, most countries in Latin America
have opted for cost-based market designs for the following reasons: 1) to ensure transparency (the dispatch and
spot prices are calculated by computer models using wellknown algorithms, with the software and system data publicly available to all market participants), 2) to guarantee
efficiency in the dispatch of hydro plants in cascade with
independent owners and multiple water uses, and 3) to
avoid potential issues with market power that could arise
in bid-based markets. When electricity markets were first
implemented in Latin America, regulators perceived that
it was important to tackle these issues to imbue investors
with the confidence to invest in new generation capacity,
which was the main goal of industry reform in these countries. In addition, regulators were concerned that the cost
of implementing a bid-based dispatch and pricing mechanism
could be prohibitively high due to the need to set up sophisticated trading platforms and market power-mitigation
mechanisms as well as educating state-owned companies
to bid rationally into these markets.
On the other hand, one of the main criticisms of the current centralized scheme used to determine water values has
to do with the sensitivity of probabilistic simulation models
ieee power & energy magazine 	

67
IEEE Power & Energy Magazine - January/February 2021

Table of Contents for the Digital Edition of IEEE Power & Energy Magazine - January/February 2021
