ELEC9715
Electricity Industry Operation and Control
Assignment 2
This assignment will be distributed to you in week 7. It is due at the end of week 9, Saturday midnight. Your submission should be a pdf document of your completed assignment submission uploaded into Moodle. The assignment must be submitted individually and must be your own work. The UNSW policy on student plagiarism can be found on the www.lc.unsw.edu.au website and you should note that we use automated software to check assignments.
The assignment will be marked out of 20 (6 marks for each question and 2 marks for overall presentation of your report). Late submission without good reason, as explained in an email to the course lecturer prior to the due submission time, will see marks reduced as per the details in the Course Guide. Late submissions must be directly emailed to the lecturer as well as uploaded into Moodle. In keeping with the recommended hours per week of study for a six unit of credit course, (9 to 10 hours of self-directed work per week) we suggest you spend around 20 hours in total on the assignment. The assignments are excellent preparation for the final exam, hence worth doing well, and it is essential that you do it yourself.
Question 1: Note that there is a spreadsheet in the assignment workbook to assist in answering this question.
Calculate the optimal unit commitment program using forward dynamic programming for the power system described below over a day’s (24 hours) operation. The power system has PV, Wind, Coal, CCGT, and OCGT generation, with operating characteristics as shown in the table below. It also has 16GWh of battery utility energy storage (BESS) with a maximum discharge / charge of 2GW (ie. an 8 hour storage system). For simplicity, assume that there are no round trip losses in operating the storage (typically around 90% round trip efficiency for li-ion). Assume that all generators can operate at any level between 0MW and their installed capacity, except that there must always be at least 1GW of coal running (minimum operating levels for the 3GW of coal plant). Note that there is a $50/tCO2 carbon price on the electricity industry. This power system might have some similarities to a future NSW generation fleet.
Generation
technology
|
Capacity GW
|
Op. cost
$/MWh
|
Emissions
tCO2/MWh
|
PV
|
6
|
0
|
0
|
Wind
|
6
|
0
|
0
|
Coal
|
3
|
30
|
1
|
CCGT
|
1
|
110
|
0.4
|
OCGT
|
4
|
170
|
0.6
|
For convenience the day is divided into 6 four time blocks. Over the day demand varies from 4 to 12GW, while the wind and solar also vary between 1-3GW and 0-5GW respectively as shown in the table below. You’ll note the daily capacity factors of around 38% and 28% respectively for this wind and solar. You can assume that the day-ahead forecasts for these renewables are very accurate. The residual demand after subtracting renewables that must be met from the thermal generation (Coal, CCGT and OCGT) is also shown. Keep in mind that 1GW minimum coal generation requirement at all times.
Hours
|
0-4
|
4-8
|
8-12
|
12-16
|
16-20
|
20-24
|
Demand
|
5
|
10
|
4
|
5
|
12
|
8
|
PV
|
0
|
0
|
4
|
5
|
1
|
0
|
Wind
|
3
|
2
|
1
|
2
|
3
|
3
|
Total PV + Wind
|
3
|
2
|
5
|
7
|
4
|
3
|
Residual demand
|
2
|
8
|
-1
|
-2
|
8
|
5
|
To solve optimal battery storage operation and mimimise industry operating costs over the day, first calculate the operating cost of the thermal generators for supplying residual demand over 4 hours in 1GW increments from 1GW (the coal power generation which cannot be turned down further) to 8GW (the maximum available thermal plant capacity). As an example I have calculated the $k cost of supplying 1GW of thermal plant (the coal unit operating at its minimum operating level) in the spreadsheet.
Now you can solve optimal battery storage operation using dynamic programming. For convenience, you can assume that the BESS is at 25% state-of-charge (4GWh of electricity) just before the first four hour block (0-4 hours). Also, the battery can only operate at 0, 1GW or 2GW charging or discharging, and stays at the chosen rate for the entirety of any four hour period. Also the BESS, and must be returned to 25% state-of-charge at the end of the last time block (20-24 hours) - otherwise, the lowest cost option would always be to empty the storage at the end of the day. You may find the table in the assignment spreadsheet of assistance in solving the dynamic programming, including the way it lays out the state space. Note that not all states at all time periods may be feasible (due to 2GW maximum charge/discharge limits on the BESS or the minimum 1GW of coal) or acceptable (you can assume that total demand must always be met). Some wind and/or solar may need to be curtailed at times as well.
What is the optimal charging / discharging trajectory for the battery storage over the day, and the lowest possible operating cost for the power system over the day? You’ll need to complete the table and then visually show the least cost BESS operating strategy over the 24 hours To assist, I have solved the State transition costs for the first time step (0-4 hours) and there is space for you to calculate the total cost of getting to each State from every other feasible or acceptable State, making it easy to then identify the least cost path to each State, and use these when calculating the least cost for the next time step.
Note that when determining the cost of meeting residual demand you need to consider reduced thermal generation (hence lower system operating costs over the four hours) if you discharge the storage, and increased thermal generation (hence higher system operating costs) if charging over that time. Also, given the maximum charging rate of 2GW the storage state-of-charge can only change by a maximum 50% (8GWh) from one four hour time period to the next.
Discuss your findings and their implications for energy storage in a power system with lots of PV, wind that mainly blows in the early morning and late evening, and demand that is relatively low overnight with a morning peak and then larger early evening peak. Do you think more BESS is required in this power system? Does the carbon price have a significant impact on BESS operation?
Question 2 Note that the assignment workbook has 30 minute NSW NEM pricing and a particular PV plant’s generation for calendar year 2024.
Your are a trader for a company that owns a 102MW utility PV plant in Western NSW that has previously sold a Power Purchase Agreement (PPA) for its entire output to a large NSW University. The agreed strike price for this PPA was $60/MWh and the 15 year contract runs from the start of 2024 through to 2039. This contract is effectively a variable volume CFD and works as follows. For every 30 minute period the spot price is subtracted from the PPA strike price and multiplied by the MWh generated during that time period to calculate the PPA value ($) over that time. Note that if the PV plant is generating at an average 100MW it generates 50MWh in 30 minutes). If the spot price is greater than PPA contract price in that 30 minutes then the generator pays the NSW University this money. If the spot price is below the PPA price then the University pays this money to the generator. The wind farm keeps all the spot market revenue.
At the start of 2025 you are asked by your company to assess the value of this PV plant, and in particular the value of it’s PPA compared to what would have happened without the contract having been signed. They provide you with the 30 minute generation (MW) of the PV plant over the year, along with the NSW wholesale spot price (see the assignment workbook) . You can assume that this PV plant was not curtailed at any time; that is, the generation trace is how much the PV plant could potentially generate in each 30 minute period. Of course the operators might have chosen a different operating strategy and turned it down occasionally.
Calculate the annual operating profit ($m/year) your PV plant would have made in 2024 considering the following scenarios:
- Your PV plant had no PPA and received only spot market revenue over the year. The PV plant is operated at its maximum possible generation at each 30 minute period.
- Your plant had no PPA and received only spot market revenue over the year. You, however, were in charge of its market participation and choose to to turn the PV plant down in any 30 minute period if you calculated that this would increase the plant’s overall operating profit.
- The PV plant does have this PPA at $60/MWh and was operated at its maximum generation potential for each 30 minute period.
- The PV plant does have this PPA but in the fine print of the contract with the University there is the following legal rider – “for the purpose of calculating the value of the CFD in each 30 minute period any negative market prices shall be defined as $0/MWh for the purpose of the calculation” The PV plant is still operated at its maximum generation potential over the year.
- The PPA has this fine print, but you have now calculated an optimal dispatch strategy to maximise the operating profit of the PV plant.
Think carefully what the optimal curtailment strategy is given both in terms of not having any PPA, or having this PPA with that fine print. It’s not necessarily the most obvious strategy :?). You can assume that your PV plant dispatch strategy doesn’t impact pricing which is a reasonable assumption. Be sure to present your results in a table showing separate CFD and spot market operating profits, as well as total operating profits for each of these 5 scenarios.
At the same time you are doing these calculations, the NSW University’s energy team is celebrating its inclusion of that fine print in its PPA contract, but is also wondering how it could manage its exposure to NSW wholesale spot prices at times when its PPA isn’t covering its demand. In particular, the university often has its peak demand of 30MW at around 6pm when the PV plant they have contracted with isn’t providing much, or in winter any, generation. You are keen to help and suggest that they might want to contract separately with a gas peaking plant for a $300/MWh cap derivative. They ask you for advice on what would be a reasonable cost ($/year) to pay for a 30MW cap covering all 8760 hours of 2025. The market expectation is that NSW wholesale market pricing in 2025 will likely be pretty similar to 2024. What option fee would you suggest to the university as a possible starting point for this negotiation with a suitable gas peaking plant? Explain how you calculate this $/year fee.
Discuss your findings and the importance of carefully designing renewable PPAs in a highly volatile spot market to manage risks for both PPA sellers and buyers, and for renewable energy projects to be responsive to changing market conditions in such cases, while PPA buyers need to consider their remaining spot exposure.
Question 3: Note there are spreadsheets in the assignment workbook which will help you lay out your answers.
A power system has a mix of old black coal plants, CCGT and OCGT gas fired generation, and wind and solar generation with overall capacity and operating costs as shown below (note the similarities with NSW). The coal plant is old and becoming unreliable with an estimated forced outage rate of 0.05 (5%) for 1500MW of the available plant (that is, you can expect that 1500MW of the coal plant is unavailable 5% of the time). The 2500MW of wind generation has the following overall probability distribution - 600MW for 70% of the time, 2000MW for 30% of the time). The 3000MW of solar generation can be roughly modelled as providing 1500MW for 50% of the time, and 0MW for the other 50% of time). Yes - these are major assumptions.
All of this serves a load with the inverted load duration curve shown below. Note that the value of any Unserved Energy (USE) is estimated to be $20,000/MWh. Assume that the wind farm output, solar farm output and load, as well as the coal plant forced outages are all completely uncorrelated (meaning that the probability of being in any of the coal, wind, solar and demand states is independent of what the other states are so yes, rather simplified). Also assume all plants can be operated anywhere between 0MW and their rated capacity, with no start-up or shutdown costs (a very big assumption for the coal plant).
The simulation period is one year ahead. Using the table provided in the assignment 2 workbook, enumerate all possible scenarios of coal generation availability, wind farm and PV output and load demand in terms of their probabilities (hence expected hours per year), ability to meet demand, and associated power system production costs ($/hr). Note that there are a total 16 possible states for this power system (2 wind states X 2 coal availability states X 2 demand states X 2 solar states). Use this table to calculate over a typical year:
(a) Loss of load probability (total hours/year)
(b) Expected unserved energy (USE) (% of load)
(c) Expected production cost over the year ($000) noting again that any USE costs $20,000/MWh.
(d) The average electricity price required to cover total production costs.
Briefly discuss your findings and their potential implications for this power system as the black coal fleet continues to age. Do you think that this complete enumeration method is useful despite the assumptions?
Let’s now consider a more realistic production cost model. The assignment workbook provides NSW 30 minute scheduled demand (MW) as well as solar (3.5GW installed), wind (2.6GW installed) and hydro (2.7GW installed) 30 minute traces for calendar year 2024. You can assume that the system has 7.2GW of black coal generation (operating cost $30/MWh) and 3GW of gas-fired generation (operating cost $150/MWh). You can assume that the wind, PV and hydro have $0/MWh operating costs..
First solve economic dispatch for this power system for each 30 minute period over the year. You can assume that the coal and gas plant have no minimum operating levels or ramp rates, and that all the gas and coal generation are always available if required. Likewise, there are no network interconnections, constraints or losses that you need to consider. For economic dispatch, you need to stack the available generation from lowest to highest operating cost up to the MW of demand for each 30 minute period. Note that Excel has a very useful function - MIN(x, y) that returns the lesser of two numbers, which just might be remaining demand vs available plant capacity for the next least cost generation technology. Any unmet demand in a 30 minute period is of course USE (unserved energy). Keep in mind also that 1000MW of any generator technology in 30 minutes represents 500MWh of generation - don’t accidentally estimate industry costs that are double the actual cost).
Hence estimate the:
(a) Loss of load probability (total hours/year)
(b) Expected unserved energy (USE) (% of load)
(c) Expected production cost over the year ($m) noting that any USE over the year costs $20000/MWh
(d) Expected generation mix (ie. the % proportion of generation over the year from each technology).
Now consider the case where 1.5GW of the coal-fired generation is increasingly unreliable with a forced outage rate of 0.05. That is, in any 30 minute period there is a 95% chance that 7.2GW of coal is available but a 5% chance of only 5.7GW being available. Estimate again the LOLP, USE, expected production cost and expected generation mix. You can use Excel’s Random number generator to determine if it is 7.2GW or 5.7GW of coal available in any and every 30 minute period. In many cases of course not all coal generation will be required so it may not change dispatch at all - at other times that missing generation may cause higher cost gas generation to run or even result in USE.
Discuss your findings, including how the results compare with the simplified complete state enumeration model used above. Again, what are the potential implications for electricity industries with growing levels of renewables and increasingly unreliable old coal-fired generation.