Leveling the Renewable Roller Coaster

As the grid transitions through 15%, 30%, 50% renewable targets and towards the holy grail of 100% renewables, the sawtooth output from renewable sources will need to be leveled into a predictable power source for use.  With higher percentages of renewables, fossil is displaced off the grid and storage of the electric energy is the proposed solution.  Storage inefficiencies reduce the overall capacity of the power generated that can be sent to the grid.  In 2017, the UK grid wind power provided 11.6% of the electric grid demand.  The capacity factors and production from Gridwatch¹ data suggests that wind generation requires 4 to 5 times the generation capacity of a fossil or nuclear plant to replace the steady state output in MWh.  To provide a stable grid with storage, up to 6 to 7 times the generation capacity may be needed depending on the storage technology used and its energy losses.     

To store and use the renewable generated electric energy, simplified graphical analysis can be done to help understand:

• the total power generated over the year
• the magnitude of the peaks and valleys of supply that occur
• the time constants of these changes
• the impact of different storage technologies that are being used to level the grid and their impact capacity available for consumption    

             The UK is being used in this study as detailed data is publicly available for analysis and review.  Gridwatch¹ data from the 2017 electrical production recorded the power output by source over the year in 5 minute intervals for 100,000+ data points. At the end of 2017, the UK had 18.9 GW of wind energy installed with 4.3 GW installed in 2017².  Approximately 1/3 of the installations are offshore that provides a steadier generation than onshore wind turbines.  The wind power generated averaged 3695 MW for the installed generation capacity of 22%. This was obtained by summing and averaging the wind power output for 100,000+ data points and dividing by the 16.75GW, the average installed wind capacity in 2017.  Output varied from a high of 9745 MW to a low of 150 MW for the entire UK wind production.  The MW output is shown in the figure below.

             Assuming that storage is 100% efficient, 3695 MW identifies steady state output that can be sustained 24/7 and 365 days a year. By sorting the data by increasing output rather than data, a relatively smooth curve can be obtained allowing an area analysis.  With fluctuating output and to store peak production, the charge capability to storage needs to be 6050 MW, withdrawal capacity to send back to the grid needs to be 3545 MW.  The green charge capacity area must be equal to or greater than the red discharge area of the storage. 

             This is a simplified calculation assuming steady state demands by the grid.  In the real world these grid demands will vary.  However, the simple analysis does provide a baseline.  Looking at grids about 50% of the peak demand from the grid will need to be delivered 24-7.  Matching the grid fluctuations to the renewable availability has proven to be difficult.     

             When storage inefficiency is added into the equation, the MWh to storage gets larger than the MHh stored. For best efficiency, it is desirable use the power generated by wind directly into the grid but that will not be feasible based on variations in production and demand and being able to supply power to the grid 24-7. 

A 50% efficient storage/withdrawal technology needs to have twice as much energy consumed in the storage process as can be recovered during the withdrawal process and lowers the total amount of power available for consumption by the grid.  Rather than having 3695MW available, storage conversion losses drops the available capacity to 3086 MW, a 17% drop in steady state capacity requiring an additional 17% more generation capacity to be installed.       

             Energy storage technologies have different efficiencies which vary from a high of 80% to 90% with battery storage to a low of 20% to 30% with hydrogen electrolysis being fed back into a simple cycle gas turbine or conventional power plant.  The UK wind data has been analyzed for storage efficiencies of 85%, 60%, 50%, 40%, and 25% to determine the wind capacity factors with the following results:

             Annual averages like the graphs and tables above help determine the investment needed to provide overall grid capacity, storage charge rates, and withdrawal from storage.  Renewable storage capacity requirements to balance load may be very short-term weather/demand events, longer term weather patterns, or more seasonal for winter to summer changes.  To review storage capacity, time related data and integration needs to be done with assumptions for the base grid output requirement and for type of storage used. 

A plot of the UK wind output for September of 2017 is presented and uses an assumed “battery storage” 85% efficiency for a baseline production of 3056 MW.  Review of the month of September confirms the annual peak charge/discharge capacity requirements of 6050/3545 MW determined are adequate with some safety margin.  The time between September 14^th and 19^th shows a period of low wind output (highlighted in red) where withdrawal from storage of 178 GWhr is needed to cover the wind output shortfall to the grid.

             Short term wind production also needs review for daily, hourly, or by the minute variations.  A plot of a 24 hour cycle of the wind output from the ISO New England website³ shows interesting patterns of wind farm output. 

• Average wind output is approximately 420 MW. 
• Total production varies over 400%
• Short term variations of 22% (138MW) occurs in 12 minutes
• A 30% (200MW) reduction occurs in an hour.
• The evening generation is quite low, which required significant natural gas and hydro production to make up the shortfall (mostly natural gas).    

             For some current cost estimations, there were 3 projects for utility scale storage bid in May 2018⁴ in Australia for battery storage.  The cost for storage varied between $520,000 to $700,000 per MWh and the cost for charge/discharge capacity varied between $1,400,000 to $2,630,000 per MW (values are in USD).  For leveling out the UK wind production, the storage costs would be approximately $93 billion USD, as the storage is a significantly the greater cost than the charge/discharge capacity.  Recent bids for “virtual power plants” made up of residential panels and storage came in at twice these budget costs.  In residential PV and storage systems, the battery costs make up only 20% of the total installed cost, so impact of the reduced battery cost on overall storage is likely not a linear relation.    

             Review of the data suggests that as the variations between peak and minimum production increases, the requirements for storage do as well bringing in storage efficiency losses and requiring higher production to make up for it.  In 2017, the UK grid wind power provided 11.6% of the electric grid demand.  The capacity factors and production from Gridwatch data suggests that wind generation requires 4 to 5 times the generation capacity of a fossil or nuclear plant to replace the steady state output in MWh.  To provide a stable grid with storage, up to 6 to 7 times the generation capacity may be needed depending on the storage technology used and its energy losses.     

References:

1. http://gridwatch.co.uk/
2. https://windeurope.org/wp-content/uploads/files/about-wind/statistics/WindEurope-Annual-Statistics-2017.pdf
3. https://www.iso-ne.com/isoexpress/web/reports/operations/-/tree/gen-fuel-mix
4. https://reneweconomy.com.au/lyon-teams-with-fluence-jera-to-pursue-big-s...