6.3. ERMUSR 06-14-2005Elk River
Municipal
13069 Orono Parkway
Elk River, MN 55330
June 8, 2005
To: Elk River Municipal Utilities Commission
Jerry Takle
John Dietz
Jim Tralle
From: Bryan Adams
Subject: Electric Resource Planning
phone: 763.441.2020
Fax: 763.441.8099
Your staff has had a number of discussions recently with Connexus and Great River
Energy concerning electric generation resource planning. How Great River Energy as
well as most of the electric industry plans for generation capacity is somewhat disturbing.
In the organization I was previously associated with, we planned from a generation
perspective for our peak day plus the 15% reserve per MAPP guidelines of which 5% of
these reserves must be spinning or on line. For example, if your projected peak load is
2000 MW, you must control or have under contract and deliverable, 2300 MW of
capacity. The 15% reserve is to ensure the utilities are doing their planning and can meet
their projected loads without gaming the other utilities. If your current load is, for
example 1000 MW, you must have 1050 MW on line and spinning. The extra 5% is to
ensure if a power plant trips off line due to electrical or mechanical failure, there is
enough spinning capacity on line to pick up the load thus preventing a major blackout.
In today's environment as generations and transmission capacity becomes tight; many of
the utilities have been hedging on the 15% reserves due to the cost and desire not to build
more power plants. Utilities are also hedging on the peak load. Instead of planning for
the peak load, Great River Energy plans for the 90% confidence level not 100% of the
possible load. Statistically, they are saying they cannot satisfy the load one day in ten
years and either the 15% reserves or market purchases will allow them to satisfy 100% of
their load. Due to the way the industry does URGE tests, 15% reserve is more like 5%.
7'he purchased power marked has also tightened up because of shortage of excess
generation with limited transmission capacity. In Great River Energy's case, 90%
confidence level at time of peak (7:00 p.m. +) is 88°F and 65 MW per degree Fahrenheit.
We have also had mild summers with no long sustained hot spells for a number of years.
ities
This planning process sounds like a receipt for disaster to me. The industry is now
saying it is acceptable to have occasional blackouts. The major blackouts in the east a
couple of years ago produced a lot of talking but little if any physical changes on how we
operate.
Attached is an interesting article titled, "Capacity Markets: A Bridge to Recovery?" that
is worth your time reading.
Capacity Markets:
A Bridge to Recovery?
A review of the ongoing evolution of market design.
BY CRAIG HART
ICAP, ICAP, UCAP, NoCAP.
Some markets have them, some
don't. Where they do exist, no
two are equal.
California was built without one, as
were the markets of Australia and New
Zealand. Northeastern U.S. markets all
have them, albeit in slightly different
variations. MISO will start one of the
biggest energy markets in the world
without one. PJM, long considered the
model market, is debating a major over-
haul to its current scheme. Others are
considering such changes as well.
As the experimentation with whole-
sale energy market deregulation contin-
ues, anevolutionary trend in capacity
market design is becoming apparent in
the United States. Most critics now
support the conclusion that en re~gy
only schemes in price-capped markets
do not provide sufficient revenues to
rn uce new generation and keep sys-
tem-critical, high-cost generation
resources in the market. As such, it is
likely only a matter of time before
energy on y mar ets are extinct.
ut, w i e it appears at capacity
markets are here to stay, there is little
consensus regarding the best design.
Markets in the United States are in a
state of flux, with debate raging over
many different capacity market pricing
schemes. The pool-wide, single-price
capacity market model utilized by PJM
and others now appears too simplistic.
New designs are calling for
ever-more complicated structures aimed
at fine-tuning the location, the timing,
and the type of generation resources
that capacity markets induce.
While the winning recipe has yet to
be selected, it is likely that participants
in certain markets will witness signifi-
cantchanges. In certain load pockets,
generators could see a meaningful
increase in gross margins if a pricing
scheme that compensates generators
based on location is adopted. Flexible
units may receive an additional bump
if changes being considered in PJM are
adopted. But, these shifts are likely only
to persist in the shoe to medium term.
While providing a much needed bridge
to market recovery For some, if the mar-
kets function properly, price differentials
should dissipate over time and bring
long-term pricing back to the mean.
Background
Capacity markets are designed to ensure
resource adequary. While various defi-
nitions exist, resource adequacy gener-
ally refers to the sufficiency of
generation resources to meet the peak
energy needs and maintain the stability
of an electric system.
In a traditional, regulated utility
world, generation plants are added after
resource planners determine the
amount of new capacity that will be
needed over the coming years to meet
demand. While they are determining
the actual amount of total installed
capacity necessary to ensure resource
adequary, resource planners also attempt
to optimize the type of capacity that is
built in terms of fuel type, technology,
and market segment (i. e., baseload,
intermediate and peaking). In this
way, the best generation mix necessary
to meet peak demand is determined
through a central planning process.
In a deregulated world, the decision
to build new capacity is made by indi-
vidual market participants instead of
through a central planning process.
The underlying question becomes how
to design a market to properly incen-
tivize profit-maximizing participants to
develop enough generation capacity to
ensure resource adequacy As the theory
goes, the answer is that the market
must provide adequate compensation
so that a new entrant is able to earn rev-
enues that are high enough to cover
costs and earn a fair return.
74vo Basic Models
There are essentially two schools of
thought when it comes to designing
markets that provide adequate revenues
to generators: energy only (referred
to as "NoCAP") and energy-plus-
capacirymarkets.'
Many have followed the NoCAP
route. In the United States, California,
ERCOT, and MISO started without an
explicit capacity market. Australia and
New Zealand, as well as the markets in
Ontario, Alberta and Scandinavia
(NORDPOOL) all have functioned
without a formal capacity support
mechanism. The general belief (hope) in
these NoCAP markets is that sufFicient
revenues will be earned from the energy
markets (plus other ancillary services
where applicable) alone to induce an
adequate amount of new generation
resources. The idea is that the inherent
volatility of the energy markets will
deliver enough high-priced hours dur-
ingsummer- orwinter-peaking periods
24 Pueuc Ununes FoprNiexnr MAV 2005 www.fortnightly.com
that a new generator would be fully
compensated and earn a fair return.
The logic for marginal, system-critical
resources is the same: The few high-
priced hours when these units are dis-
patchedwill provide enough compensa-
tion to cover going-forward costs and
keep the units from retiring.
The other option, of course, is to
couple a capacity mazket with the
energy market. PJM, ISO New Eng-
land, the New York ISO (NY--ISO), and
others have used'vaziants of this theme
fairly successfully. Energy-plus-capacity
markets generally function by compen-
sating generators based on total
installed, or available, capaciryZ Genera-
tors receive a payment that supplements
revenues earned in the other markets.
The idea is that total revenues should
be sufficient to cover going-forward
costs at a level that is compazable to
what generators eazn from hourly dis-
patch in an energy-only market.
Bumps fn the Road
As deregulated markets collect a histori-
cal record that includes the experience
of a full boom-bust business rycle, sev-
eral shortcomings in existing markets
are becoming apparent. For NoCAP
markets, it is the institution of price
caps that is arguably the most notable
contributor to their inability to provide
a stable and sufficient revenue stream to
generators. Others with capacity maz-
kets in place are finding that simply
adding generic capacity resources some-
wherewithin the market footprint may
not be the optimal path to ensuring
resource adequacy What started as sim-
ply aquestion of how much, has re-
cently matured to a question of what,
where, and when.
Price Caps. In U.S. energy markets,
extreme price volatility is unpalatable to
the general public. As a reaction to high
prices in the Midwest in 1998, PJM in
1999, and California in 2000, many
www.fortnightly.com
markets created price ceilings (or price
caps) to protect the public from extreme
energy price volatility. But the problem
with this type of protection is that it sac-
rifices the prospect oflong-term market
stability for the sake ofavoiding adam-
agingsummer newspaper headline.
This is especially true in NoCAP
markets. Marginal generators in
NoCAP markets are dependent on the
few days when the intrinsically volatile
energy mazkets deliver high-priced
hours. The ability to capture revenues
during these high jpriced times gives
mazginal generators the ability to cap-
ture a sttfficient level of revenues. But,
with price caps in place, generators lose
this revenue which tends to dampen the
new build signal and undermine the
ability of system-critical resources to
capture adequate revenues.
Location. Many have begun to rec-
ognize that the regional diversity of the
geographic footprint covered by power
pools is such that they need to worry
about where generators are encouraged
to btuld. Given the constraints of the
transmission system, it is critical that
new generation be located as close to
load as possible. For example, building
new generation in western Massachu-
setts does not solve the supply situation
in southwestern Connecticut.
In single-price capacity markets,
there is no incentive or price signal pro-
vided by the capacity market that
encourages new generation to build in
an area that benefits the grid. Differ-
ences in locational marginal pricing
(LMP) encourage generators to build
closer to higher-priced locations, but
this does not provide enough benefit to
marginal generators.
.The shortcoming ofnon-location-
allybased capacity markets can be wit-
nessed in part by the need for many
markets to continue to maintain relia-
bilirymust run (RMR) programs.
RMR payments are designe to com-
pensate system-critical resources that
are needed to maintain system reliabil-
ity. These resources are often older units
that run infrequently due to inefficient,
out-moded equipment with high heat
races. In many cases, these units would
not cover their going-forward costs and
would be retired if they didn't receive
additional monetary support From
RMR payments.
For example, units like El Segundo on
the beach in Los Angeles or Devon on
the coast of Connecticut are needed to
maintain system reliability. In those areas
it is very difficult to site new transmission
lines, and probably impossible to site a
new power plant anywhere near El
Segundo, so the system needs to ensure
that these units earn sufficient revenues
to keep them from retiring. V~'ithout a
capacity market that provides additional
compensation for these generators (or
more modern substitutes), the power
pool must continue to prop them up
with administrative RMR payments.
Operating (and Fuel) Diversity. In
addition to issues surrounding location,
questions have been raised about the
ability of capacity markets to address
the operating profile of new equip-
ment. The argument is that even if you
get enough capacity at the right place,
you still need to worry about the type
of equipment that is built. The system
needs resources that have operating
flexibility and also contribute to the
fuel diversity of the overall fleet.
The "dash-for-gas" witnessed in the
recent past created a glut of homoge-
nous units. One problem that has sur-
faced as a result of this monolithic block
of capacity is that supply is very depend-
ent on gas prices. Given the correlation
of gas prices to heating needs in winter,
regions have seen shortages and price
spikes when gas-fired generators sold gas
into the heating market rather then gen-
erating electricity. Inaddition, many of
the units have similar operating chazac-
Mav 2005 Pueuc Ununes Foerxiexnr 25
~il~!I
/".
6
~~h'
1!'I
teristics. Many believe that capacity
markets need to have the ability to rec-
ognizefuel and operating diversity to
avoid dependenry on one fuel type and
to maintain a fleet of resources that is
capable of meeting all of the needs of the
system operator in terms of ability to
start quickly and follow load.
Demand Curve. Finally, there has
been one other major shortcoming
observed in capacity market design that
occurs on the demand side of the equa-
tion, asopposed to the supply side
issues like those already discussed. Most
markets have utilized what is referred to
as a "vertical demand curve" to date.
This term refers to the construction of
most markets where demand is set
equal to the target reserve margin
which, in effect, creates a fixed quantity
that is completely price insensitive.
While a description of the full
dynamic of the pricing structures of
demand curves are beyond the scope of
this article, the vertical demand curve
typically has created very volatile pric-
ing. Historical pricing patterns in exist-
ing capacity markets have tended to be
shaped by boom/bust or floor/ceiling
behavior where prices hover near zero
when the market is oversupplied and
jump to the capped price or deficienry
level when the market passes through
equilibrium and enters a period of
shortage. This binary pricing pattern
produces unwanted volatility that
reduces the predictability of revenues
for suppliers. Reduced predictability
undermines new entrant planning or
forces builders to raise costs by includ-
ing apremium to cover this price risk.
Capacity Market Evolution
In response to these perceived short-
comings, many markets are consider-
ing, or already are undergoing, massive
overhauls. Some are adding capacity
markets for the first time, while others
are adding new features to existing mar-
kets. These overhauls are~spurring a new
round of thinking on calpacity market
design that appears to be leading to
interesting-and increasjngly com-
plex-new designs. While the final
direction of the markers t•emalns to be
seen, what is becoming a,~parent is a
natural evolution in cap city market
form and function. I
The bottom of the evdlutionary lad-
der (see Figure 1) can be defined as begin-
ningwith NoCAP markets as seen in
California and ERCOT. Given the issues
associated with price caps in most mar-
kets, this structure appears to be insuffi-
cient. From there we move up the evolu-
tionaryscale to simple, pool-wide capac-
itymarkets like the current form of the
PJM installed (ICAP), or unforced
26 Pueuc Ununes Foeneaxnr Mnv 2005
(UCAP), market. But, as discussed
above, this relatively simple structure
addresses the question ofvolume only.
The locational ICAP market
(LICAP) structure that follows the rela-
tivelysimplistic ICAP/UCAP model
adds a locational component to the
pricing structure to add criteria for
determining where in the pool new
resources are constructed. The pro-
posed system in New England, as well
as the current system in New York, are
examples of this LICAP approach.
In addition to addressing location,
the LICAP phase also has included
modifications to the market-demand
curve. By constructing what is referred
to as a "sloped" or "curved" demand
curve, market designers are attempting
to fine-tune the timing of the new build
signal and eliminate the binary pricing
behavior observed in markets with verti-
cal-demand curves. The sloped-demand
curve attempts to add subtlety to the
market signals by producing increas-
ingly high levels of compensation for
generators when the market moves
towards scarcity levels, and by gradually
decreasing levels of compensation when
resources are adequate.
PJM now is considering taking this
a step further, to what can be referred
to as "LICAP Plus." If the original PJM
structure simply addressed the issue of
how much, and the New York/New
England LICAP markets add function-
ality for addressing when and where,
the new PJM design also will help
determine what type. The scheme
under discussion in PJM is designed
to address the operational flexibility
of the resources that are built. The
proposal includes criteria to provide
price distinction based on the ability
of resources to follow load or provide
quick-start functionality. This will
help ensure that the generation mix
as a whole is diverse enough to
provide the full range of resources
www.fortnightly.com
needed to meet system needs.
So what does all this mean? In terms
of market evolution, there is no reason to
expect that all of these steps will be fol-
lowed inorder byall markets. As experi-
encebuilds and historical examples accu-
mulate, it is likely that markets will level-
jumpfrom NoCAP to LICAP, for in-
stance, as is being considered in Califor-
nia (see Table 1). Certainly, if a LICAP
Plus market appears to work, we may see
markets open in the future with that
structure used from the beginning. Re-
gardless, the current state of the various
markets in the United States provides a
live example of the progression of
thought on how best to ensure resource
adequary in deregulated markets.
The outcome of the discussions
under way in PJM will have a significant
impact on the future form of capacity
markets. PJM has been viewed for some
time as a model market, so others likely
will continue to follow its lead and
adopt similar market structures. In addi-
tion, MISO is under order from the
Federal Energy Regulatory Commission
to produce a system that is consistent
with PJM. Given these developments,
within several years the Northeast likely
will be dominated by LICAP markets,
while LICAP Plus could stretch from
the Mid-Atlantic to Montana. Develop-
ments in California and ERCOT
should add additional momentum to
the LICAP model in future years.
For market participants, these new
structures could have material impacts,
at least in the short term. For a generator
in Connecticut, for example, the addi-
tion of LICAP could mean the differ-
encebetween less than $1 /kW-mo. in
capacity compensation or $5 or $6/kW-
mo. For a generic 600-MW combined-
rycleplant, this could add $40 million
per year in revenues. To put this in per-
spective, anadditional $40 million
would increase gross margins from 15 to
40 percent for this typical plant, which
I
Market Current Future Comment
Capacity Market Capacity Market
CA None (energy-only) ICAP/LICAP Market Design 2002 (MD02) includes an Available
Capacity (ACAP) market that will function like ICAP/
UCAP markets. Also considering adding a LICAP
component.
ERCOT None (energy-only) Under Discussion Considering various capacity market options as part
of overall market overhaul.
MISO None (energy-only) LICAP Plus Under FERC order to design a resource adequacy
scheme that is consistent with current (and future)
PJM system. Started market April 7 without capacity
market
NY ISO LICAP` No Change New York is a three-zone LICAP market with demand
curve. Pricing zones for New York City, Long Island
and Rest of State.
ISO NE ICAP' LICAP Moving to a 5 zone LICAP market with demand curve,
Wiil look a lot like New York.
PJM ICAP' LICAP Plus In prtx~ss of designing a new Reliability Pricing Model
that will include price adders for location and operating
flexibility. Also considering adding a demand curve.
1. Markets calculated based on unforced capacity availability which adJusts available capacity to reflect forced outages.
Referred to as ICAP markets that utll¢e UCAP calculation methodology.
goes a long way toward covering debt
service on struggling merchant plants.]
For plants in other parts of NEPOOL,
the shift will be less dramatic or perhaps
even nonexistent. A plant in the pro-
posed "Rest-of--Pool" zone will see little,
if any, price appreciation.
While price shifts in tight locations
like southwest Connecticut and eastern
PJM may help struggling generators in
the short term, they are unlikely to per-
sist over the long term. There are essen-
tially three outcomes for the revised
markets, all of which should lead to
prices that trend toward the mean.
First, if the markets function as
intended, high prices in load pockets
will attract new resources that will alle-
viate the shortage and bring the zone
back into equilibrium. This will serve
to drive prices down in this region over
a relatively short time horizon.
Another possibility is that new
transmission capacity will be built that
alleviates constraints within a zone and
allows for more import capacity. This
outcome also would put the same
downward pressure on prices. In fact,
some are calling already for NEMA/
SWCT zonal prices in New England to
converge on Rest-of--Pool prices within
one year of LICAP star because of
planned transmission upgrades. Finally,
if a new market structure leads to sus-
tained high prices for merchant genera-
tors, it will be apparent that the market
is not working as intended and regula-
tors will be forced to reconsider the
market structure due to the impact on
consumers.
Craig Hart is vice president with US Power
Generating Co., established fo acquire and
manage merchant power projects. He can
be reached at chart@uspowergen.com.
Endnotes
1. This categorization excludes other ancillary
services including spinning reserves, regulation,
black-start capability, etc. Nevertheless, it is
important to note that it is the combination of
energy, ancillary services, and capacity products
that provides a total revenue package to gener-
ators. But, since these other ancillary services
do not differentiate the individual resource ade-
quacy markets, a discussion of them is consid-
ered out of scope of this article.
2. Many markets account for scheduled out-
ages in deterrrtining unit availability.
3. Final market designs may include mecha-
nisms for subtracting infra-marginal revenues
when determining capacity compensation. If
this is ultimately adopted, assuming all else
remains equal, the more a generator earns in
the energy market the less it will earn in the
capacity market.
;,
~,
www.fortnightly.com Mnr 2005 IMeuc Ihuntes Foettuextiv 27