XYLENE POWER LTD.

DISTRIBUTED ELECTRICITY GENERATION

By C. Rhodes

INTRODUCTION:
Ontario needs major private sector investment in non-fossil fuel Distributed electricity Generation, natural gas co-generation and Energy Storage to meet the electricity supply shortage and the required reduction in CO2 emissions that are anticipated during the next decade.

ELECTRICITY GENERATION PROBLEMS:
Most of Ontario's electricity is presently generated by conventional hydro-electric dams or by large thermal-electric generating stations. Most of the rivers with sites that are suitable for large hydro-electric dams have already been harnessed. Fossil fuel thermal-electric generators are a problem because they are large sources of the man made CO2 that causes global warming. Nuclear thermal-electric generators have proven to be quite expensive, need to be base loaded, and have complex safety and waste management issues. All thermal-electric generators reject large amounts of heat that significantly alters the local environment. In order to obtain electricity without the aforementioned problems attempts are being made to use distributed renewable electricity generation.

DISTRIBUTED ELECTRICITY GENERATION:
Distributed electricity generation involves taking advantage of economic and environmentally acceptable sources of energy for electricity generation wherever and whenever they occur. Unconstrained distributed electricity generation has an output power versus time profile that is determined by available sun light, available (wind velocity)^3, available river flow, and by the rate of use of low grade heat for purposes other than electricity generation. Intermittent unconstrained generation triggers a requirement for additional transmission.

RELIABILITY OF DISTRIBUTED ELECTRICITY GENERATION:
The reliability of an electricity system with distributed generation rests on there being a large number of statistically independent distributed generators that collectively provide sufficient average power to meet both the local area grid load and the energy storage system losses. On-going voltage regulation requires control of energy storage and sheddable load as well as some generation constraint. To minimize dependence on long distance transmission local generation should be operated to approximately equal local load. The process of matching local generation to local load requires excess generation capacity and controlled generation constriant.

GENERATION CONSTRAINT:
As intermittent unconstrained generation is added to the electricity system balancing sheddable load, energy storage and constrained generation should also be added to the electricity system. In the absence of these balancing components the distributed generation must be constrained to achieve load following. This generation constraint reduces each generator's average power output by about 50%.

The generation constraint requirement can be met by requiring all distributed generators to constrain their outputs according to a specified negative slope % of available power versus voltage curve. However, implementation of this generation constraint methodology requires acceptance by the parties of appropriate contract terms.

Operation of unconstrained intermittent generation leads to loss of energy revenue by load following generators. This is a financially intolerable state of affairs that has yet to be adequately addressed by the Ontario Power Authority (OPA).

PROFILE MATCHING:
Congestion Factor based electricity rates should be used to encourage use of energy storage for leveling generation and load profiles. To the extent that energy storage does not achieve generation and load profile matching, then generation and/or load under dispatch control must be used to achieve the required profile matching. The economics of distributed generation depend on how well the total seasonal distributed generation output matches the grid load profile. The cost of seasonal energy storage is quite large.

GENERATOR INDEPENDENCE:
An issue that has not been adequately addressed by the OPA is requiring distributed generators to be self excited to reduce their dependence on other generation for voltage regulation, reactive power and black start.

DISTRIBUTED GENERATION TYPES:
Two significant sources of distributed generation energy are wind and co-generation. It is helpful to examine the costs of both types of distributed generation. The mathematical development for other forms of distributed generation is analogous.

WIND ENERGY:
The OPA has determined that the value of unconstrained land based wind energy is about $.15 / kWh. Of this amount $.135 / kWh is the base rate paid by the OPA, $.01 / kwh comes from the federal government and about $.005 / kWh comes from an inflation adjustment to the OPA payment.

When statistically independent unconstrained distributed generation is constrained to follow the actual Ontario load profile the average generator output is reduced by a factor of two. Hence the average cost of wind energy increases to:
2 X $.15 / kWh = $.30 / kWh

The transmission/distribution and regulatory costs are additional.

The power output from a wind generator is proportional to (wind velocity)^3. Doubling the wind velocity increases the power output by a factor of 8. Hence the available wind energy is a strong function of the wind velocity distribution over time.

A fundamental issue with wind generation in Ontario is that the average power output in the winter is about twice the average power output in the summer. In the summer there can be periods as long as ten days when the wind velocity is low and almost all the energy requirements must be met from energy storage. In the winter there can be periods as long as five days when the wind velocity is low and almost all the energy requirements must be met from energy storage. The experimental data published in Section 6.1 of the Ontario Wind Integration Study shows that the minimum energy storage required to smooth out week to week variations in wind output is about 480 kWh for a 10 kW peak output wind generator.

WIND GENERATION ENERGY STORAGE SYSTEM ASSUMPTIONS:
1. A representative unconstrained wind generator has a peak output of 10 kW, a winter month average output of 4 kW and a summer month average output of 2 kW.
2. The adjacent energy storage system is rated for a maximum input of 8 kW and a maximum output of 4 kW.
3. The amount of energy storage required to smooth the output of a 10 kW peak output wind generator is 480 kWh.

A major component of the energy storage system cost is the cost of the chemicals used to store the energy. Assume that the chemical system used for energy storage is sodium-sulfur-nickel chloride (Na-S-NiCl2).The characteristics of this electrochemical storage are set out at: Na-S-NiCl2 Electro-Chemical Energy Storage.

The cost of the chemical inventory is about:
$25.00 / kWh X 480 kWh = $12,000
The chemical inventory could be contained in a battery casing assembly worth another $2000 and associated with a 6 kW inverter worth $6000, for a total energy storage system cost of:
$12000 + $2000 + $6000 = $20,000

ENERGY STORAGE LOSSES:
Assume that 60% of the energy input to storage is recoverable.

Let G = average wind generator output

Assume that half the unconstrained wind generated energy goes into storage.

Then the net energy output is:
G(0.5 + (.6)0.5) = .80 G

Hence the storage energy loss is:
(1 - .80)G = .20 G

Hence if an unconstrained wind generator has an average output of 0.30 X the peak output, with the storage system the average output will decrease to:
0.30 X 0.80 = 0.24 X the peak output.

The annual net electricity output from the combined wind generator and energy storage system is:
10 kW X 8766 hour X .24 = 21,038.4 kWh / year

The energy storage capital cost / kWh / year is:
$20,000 / (21,038.4 kWh / year) = $.9506 year / kWh
With 7.34 year amortization this capital cost increases the cost of wind generation by:
($.9506 year / kWh) / 7.34 year = $.1295 / kwh
The cost of the unconstrained wind generated electrical energy is:
$.145 X (0.30 / 0.24) = $.18125/ kWh
Hence the total price of winter weighted load following electicity from wind generation with about a week of energy storage is:
$.18125 / kwh + $.1295 / kWh = $.3107 / kwh ~ $.31 / kWh

The cost of storage / annual average kW is:
$20,000 / 2.4 kw = $8333.33 / kW

The representative wind farm projects studied by the OPA indicate that the present cost of wind generation without energy storage is about $9000 / annual average kW
so the total capital cost of reliable wind power is about:
$8,333 / kW + $9000 / kW = $17,333 / annual average kW output assuming a winter weighted market.

The comparable cost for nuclear electricity is about $12,000 / kW assuming a non-seasonal market and neglecting the cost of energy storage for load following.

To follow the daily load variation the nuclear system can be constrained by .674, so that its cost per average kW becomes:
($12000 / kW) / .674 = $17,804 / average kW output

Hence the cost of nuclear generation with load following constraint is similar to the cost of wind generation with load following energy storage. However, nuclear generation has the advantage that it better meets the peak load in the summer. Wind generation with energy storage has the long term advantage that it better meets the peak load in the winter.

SUMMARY:
Under the aforementioned costing assumptions, on a per average kW basis the costs of nuclear generation and the costs of wind generation are similar, and the choice between them should be made based on the shape of the seasonal monthly average electricity load profile. A flat load profile points to nuclear generation. A load profile with winter consumption twice summer consumption points to wind generation. Anything in between those two extremes indicates that a mix of nuclear and wind generation is optimal.

The table in Equipment Financing shows that with government guaranteed debt financing the cost of electricity from output constrained nuclear generation that costs $12000 / kW and does not have energy storage but is constrained for load following is:
Vg = $.4878 / kWh + transmission-distribution cost

Hence the cost of wind generation and associated energy storage with government guaranteed debt financing will likely be about:
($17,333 / $17,804) X $.4878 / kWh = $.4749 / kWh + transmission-distribution cost.

In the absence of seasonal pumped hydraulic energy storage wind energy only has an economic advantage over nuclear energy when serving the portion of the total electricity load that is winter weighted. However, monthly energy consumption data from the IESO indicates that as of 2006 the portion of the total electricity load that was winter weighted was zero. IESO data from its generator reports indicates that during 2006 the monthly generation in Ontario in TWh was:
Jan: 13.93 TWh
Feb: 13.15 TWh
Mar: 13.75 TWh
Apr: 12.40 TWh
May: 12.49 TWh
Jun: 12.78 TWh
Jul: 14.20 TWh
Aug: 14.02 TWh
Sep: 11.96 TWh
Oct: 12.49 TWh
Nov: 11.88 TWh
Dec: 13.05 TWh

The total generation in 2006 was 156.1 TWh.

In order for the electricity consumption in Ontario to become winter weighted there must be an increase in electricity used for water heating and space heating. Such an increase could be triggered by a sustained increase in the price of fuel oil. In July 2008 the retail price of fuel oil reached $1.40 / litre plus GST without any fossil carbon emissions tax. According to NRCAN in 2006 Ontario consumed 1,068 million litres of fuel oil. The heat that is available from this fuel oil burned at 85% efficiency is:
.85 X 38.2 MJ / lit X 1,068 X 10^6 lit/year = 34.678 X 10^9 MJ / year
=34.678 X 10^9 MJ / year X 10^6 J / MJ X 1W-s/ J X 1 kW / 1000 W X 1h / 3600 s
= 9.6327 X 10^9 kWh /year
= 9.6327 TWh / year

Viewed as a fraction of total generation in Ontario this opportunity for economic wind power is about:
9.6327 TWh / 156.1 TWh = .0617 = 6.17%
This is a much smaller fraction of total generation than many wind power advocates have claimed. In order for this fraction to significantly increase the price of natural gas must approach the price of oil. Since this fraction is a relatively small part of total generation it is necessary for the government to raise the present 15,000 MW cap on nuclear generation. Alternatively it is necessary to build pumped hydraulic storage sufficient for seasonal storage of wind energy.

A 10 kW peak output wind turbine with a capacity factor of 0.3 and an effective energy storage efficiency of 0.6, as described above, has an annual net output of:
10 kW/turbine x 8766 h/year X .3 X .80 = 21038.4 kWh / year-turbine
Thus the maximum number of such 10 kW wind turbines that the Ontario energy system could ultimately accept is:
(9.6327 X 10^9 kWh / year) /(21,038.4 kWh / year-turbine)
= .45786 X 10^6 10 kW wind turbines
= .458 X 10^4 1 MW wind turbines
= 4580 1 MW wind turbines
= 2290 2 MW wind turbines

If this conversion is spread over a 10 year period the annual construction rate would be:
2290 turbines / 10 years = 229 2 MW turbines per year.
This construction rate is well within industry capability.

NATURAL GAS CO-GENERATION:
This author's practical experience during the period 1995 to 1999 indicated that a budget of about $200,000 was required to design, supply and install a 100 kW natural gas cogeneration retrofit in a major building. Since that time the cost has probably increased to about $250,000 for similar work. Assume that there is a TOU electricity rate with 12 h on-peak and 12 h off-peak every day. As shown on the web page titled "Equipment Financing", financing this equipment to run at a 50% load factor with private funds requires an average electricity price Vp while the equipment is operating of:
.48 X ($2500 / kW) = (8760 h X .5) X (Vp - F)
or
Vp - F = (.48 X $2500) / (4380 kWh)
= $.274 / kWh,
where F = fuel and consumeable costs.

For thermal load following CHP systems (a behind the meter distributed generation opportunity for multi-residential buildings), for every kWh of incremental electricity generation a corresponding additional kWh of heat must be provided to the generator's prime mover. In most cases the fuel is natural gas which is burned at an efficiency of about 80%. The cost of the incremental extra heat which is converted to mechanical and then electrical energy is about:
$.426 / m^3 X 1 m^3 / 10.4 kWh X 1 / .8 = $.0512 / kWh

Since natural gas is a fossil fuel it will attract an equivalent fossil carbon emissions tax. The webpage titled Fossil Carbon Emissions Tax shows that with a fossil carbon emissions tax of $200 / tonne CO2 this cost of natural gas will be:
$.0512 / kWh X (.09072 / .0486) = $.0956 / kWh

For natural gas fuelled systems there is also the issue of depreciation of the prime mover. The prime mover may be a reciprocating engine that at 100 kW costs about $32,000 to replace and that lasts 16,000 hours or may be a recuperated combustion turbine that costs $64,000 to replace but that lasts 32,000 hours. In either case the prime mover has depreciation costs of:
$32,000 / (100 kW X 16,000 H) = $.02 / kWh

There is also a problem of consequential damages. That is, a prime mover failure may damage a generator, or a generator failure may damage a prime mover, or an electrical/electronic failure may damage mechanical equipment. When consequential damages, oil, filters, etc. are taken into consideration it is prudent to use a depreciation cost of $.03 / kWh.

Hence a co-generation system only makes economic sense for a building owner if he can operate at full power over the 12 hour daily load peak and if he receives an average price for the co-generated electricity energy of:
$.274 / kWh + $.0956 / kWh + $.03 / kWh = $.40 / kWh
plus any applicable transmission-distribution costs.

This rate can be slightly mitigated via CCA Class 43.2. Generally the Ontario government has no interest in purchasing or being involved in financing equipment that is as maintenance intensive and building integral as a co-generation system. In addition, a co-generation proponent needs long term price protection against increases in the cost of natural gas.

Thus if and when there is an electricity energy rate offered to building owners that exceeds $.40 / kWh + transmission-distribution cost for 50% of the time it may be beneficial for a building owner to operate behind the meter natural gas fuelled electricity generation during the on-peak period.

THE DEMAND METERING ISSUE AND ITS SOLUTION:
A significant problem affecting co-generation is that the existing OEB approved transmission/distribution rates for large buildings (over 50 kW peak demand) act as a major disincentive for behind the meter generation in these buildings. In most large buildings it is physically impractical or cost prohibitive to connect a generator directly to the grid. It is much more practical to connect internal generators behind the LDC electricity meter. The output of such generators has the effect of reducing the number of kWh registered by the LDC electricity meter.

The problem is that, with present Ontario Energy Board (OEB) approved Local Distribution Company (LDC) electricity rates, when a behind the meter generator is shut down for routine service or, in the case of CHP, for lack of heat load the building's peak kW or peak kVA meter registers a peak, causing the building owner to be billed for extra kW or kVA equal to the entire generator capacity. This extra billing makes behind the meter generation in large buildings less economic. The long term solution to this problem is to change the electricity rate structure so that the building owner is billed for transmission-distribution in accordance with the daily congestion factor. This new billing methodology is simple to implement with an electronic meter. This methodology is fair to the LDC, and provides the building owner with financial incentives to keep his power factor high, his harmonic distortion low and to flatten his load profile.

However, it may take many years to implement these rate structure and metering changes via the existing Ontario Energy Board (OEB) approval process. During the transition period it is contemplated that LDCs will continue to bill customers using existing meters. Customers that implement internal self generation measures should be given priority for supply and installation of new electronic meters. These new electricity meters must measure and separately cumulate the number of kWh imported by the customer and and the number of kWh exported by the customer. Since the rate per kWh for import from the grid is higher than the rate per kWh for export to the grid, customers increase their electricity costs if they present time varying and reactive loads to the grid.

CONCLUSION:
The biggest issue facing electricity regulators in Ontario is that the cost of new non-fossil fuel energy is about five times larger than the present Hourly Ontario Electricity Price per kWh (HOEP). Reducing this energy cost requires flattening the provincial power versus time electricity load profile. Until the Ontario Energy Board seriously addresses this issue the people of Ontario will become even more dependent on fossil fuelled electricity generation for load following. This author sees no solution other than to:
1. Apply a fossil carbon emissions tax of about $200 per emitted CO2 tonne to fossil fuel electricity generation. Use the revenue from this tax to reduce the stranded electricity debt principal, thus allowing future use of government financing guarantees to reduce the cost of major electricity projects;
2. Implement a new system for electricity energy and transmission/distribution cost allocation based on direction sensitive interval kWh metering and daily congestion factor. The objective of this new cost allocation system is to cause sufficient hour-by-hour electricity price swing to cause both generators and loads to flatten their net power versus time profiles.
3. Implement electricity end user pricing with sufficient seasonal price swing to finance central seasonal energy storage.

This web page last updated December 16, 2009.