Post

The Coming Tricky Peak Load Wave of Electric Vehicles – And What to do About It

Posted to GridIntellect, LLC – A Veteran-owned Company in the Load Management Group
image credit: Image provided by Alphabet Fleet Services Sweden, a BMW Group Company
Stuart McCafferty's picture
President GridIntellect

Curious researcher, technologist, distributed energy expert, and climate change believer.  Genuinely focused on doing good for the industry and the future.

  • Member since 2018
  • 91 items added with 65,161 views
  • Mar 19, 2021 2:33 pm GMT
  • 947 views

THE COMING TRICKY PEAK LOAD WAVE OF ELECTRIC VEHICLES

By Eamonn McCormickStuart McCafferty, & Hans-Arild Bredesen

NEWS FLASH!  Today’s grid was not designed for today’s electrified, intermittent generation, green ecosystem – and meeting tomorrow’s electricity needs will be even more challenging!  Ok, so maybe that isn’t a surprise to anyone anywhere.  As the world becomes more and more dependent on reliable, inexpensive, resilient electric power, our electric power grid and those who operate it have become central to meeting national and global economic and societal needs.  Business and societal expectations for the grid are that it operates at near perfection.  Power outages and higher electricity costs make national news whenever they occur with finger-pointing, anger, and indignation directed towards the grid and market operators, some of it rightly so.

If you stand back and think about the way that we plan for grid upgrades, it becomes very clear why DSO and TSOs are struggling to meet rapidly growing customer expectations and electricity demand needs.  Our industry plans for PEAK LOAD needs that occur only a few times a year.  So, we overbuild the infrastructure for most times of the year.  That’s not necessarily a horrible idea and it has worked reasonably well in the past, but as we look forward to high penetration electric vehicles and lower reliance on fossil fuels for EVERYTHING, the amount of planned electricity capacity to meet peak load needs of the near future are mind-boggling AND EXPENSIVE!

So, just think about what’s coming next with a huge wave of completely unfamiliar and untested electricity customers – electric vehicles. These guys are different.  First of all, they are mobile and could literally show up anywhere anytime on any network.  Second, they are not traditional “rate payers” (a derogatory term I believe we should abandon).  They are intelligent customers that can consume, supply, and store electricity – ok, we can call them prosumers (another unsatisfying name), but consider them as grid actors that can actively participate and coordinate with for grid services and market participation.  Third, when electric vehicles (and their charging stations) are aggregated together, they can have significant negative impacts to the grid if they are not coordinated with properly.  Frankly, electric vehicles deserve their own customer category, especially when grouped in fleets since they will be critical pieces of the overall grid operations and will behave differently than residential and C&I customer premises – remember, you heard it here first!

Regardless, the tidal wave of EVs is coming.  In 2016 Rocky Mountain Institute (RMI) estimated that a 23% penetration rate of EVs in California with uncontrolled charging could increase peak load by 11%.  That could very well be an underestimation for the effect on peak load and is almost certainly a very conservative estimate for EV penetration over the coming decade (there are already approximately 350k EVs in California today).  Per RMI, "Given 1.1 million EVs in service by 2030, our model estimates that the representative utility will need to make cumulative transmission and distribution investments of $2.8 billion through 2030, for an estimated grid capacity upgrade cost of $2,600 per EV. . . Most of these costs are from investments in distribution assets; transmission assets account for only $110 per EV in costs or less than 5% of the total investment costs."  Some interesting points made by RMI – that these are enormous investments and the vast majority of the money spent will be on the distribution networks, not transmission.  The California Public Utilities Commission (CPUC) has a much more aggressive goal of 5M EVs by 2030, meaning about 1/3 of all registered vehicles in California.  The model we developed below shows the CPUC’s goal with an annualized forecast going to 2030.  Regardless of which model is most accurate, both show a significant increase in the EV load that the grid will need to serve.  Of course, this will place even more demands on the grid due to transportation electrification, as there will be mass transportation and fleet charging stations that are not part of the calculus of our modeling.

Linear Projection Assuming 33% EV Penetration by 2030 (per CPUC goals)

As a utility, the dramatic increase in electric vehicles and the subsequent electricity demand must make you nervously smile.  Southern Company’s Smart Neighborhood pilot results showed that EV charging could make up 15-20% of home energy usage.  The use and dependence on electricity is only going to increase and provide “job security”.  But, managing all the new grid variables in a coordinated and reliable way is daunting, especially with today’s centralized architecture and approach.  This is not a top-down, transmission-centric solution anymore.  The grid management equation BEGINS AT THE CUSTOMER.  It’s just the opposite of the way it is done today.

As a customer, the idea of a grid that must provide much more capacity than currently available is not all that great.  We are already experiencing the vulnerabilities of the grid to weather, fires, and unbalanced supply and demand.  The customer’s increased dependence on electricity and the additional potential supply issues due to more EVs, too much sun, too little sun, not enough storage, etc. raises more concerns for reliable, resilient electricity.  What this means for “rate payers” (sorry) is simply higher rates and more reliance on a monopolistic industry with little or no customer choice on which provider to choose.  Even so, the costs of implementing the necessary grid upgrades would likely be the same or less than the customer’s costs of purchasing petroleum fuels for transportation, and would certainly have other environmental and social benefits.  Still, the customer’s local electricity provider effectively becomes the single point of failure for all energy needs, and given some of the recent high visibility outages, it is not an ideal situation.

But, making expensive grid upgrades to increase peak load capacity is not the ONLY choice.  Keep reading.

PEAK LOAD AND DEMAND CHARGES

Peak load is a significant and costly issue facing the electricity industry. Clean electricity is the only pathway to a clean power grid, clean transportation, clean heating and cooling, clean lighting, and a decarbonized society. As we decarbonize, it adds additional strain on an already overstretched grid.

Without question, the biggest issue is peak load. So what do we mean by peak load? According to EnergyWatch, we can "Think of consumption (kWh) like a car's odometer, tracking the total amount of miles driven, and demand (kW) is like the speedometer, tracking the highest speed you've traveled at a particular point in time.  This highest speed you've traveled, your peak kW demand, affects both the delivery and supply costs you pay."

For a typical customer in a wholesale market, this peak kW, or peak load contribution/capacity tag/cap obligation (various names depending on market), can comprise up to 50% of your supply bill and more than 30% of your total electricity bill (supply + delivery) due to “Demand Charges”. Therefore managing peak load for both customers and the grid will be of paramount importance.

So why is this such a big problem, and why should you care? For most utility territories, C&I customers (generally accounts with a peak kW >300) are billed both kWh and kW Demand Charges (as opposed to residential, which is just based on kWh consumed).  The delivery Demand Charge you're assessed each month is based on the peak kW you register during that billing period, even if it's just for a 30-minute interval.  Imagine a $20k normal monthly bill suddenly becomes $26k due to a 15 or 30 minute load overage.  It can have a dramatic effect on bottom lines for companies that experience periodic Demand Charges.  Numerous companies have sprung up to help businesses manage their load and avoid these sometimes surprising Demand Charge costs. 

Supply and demand are two sides of the same coin.  You can theoretically manage both, so upgrading the grid to support higher supply capacities is not the only answer.  If the load can be managed to reduce load or “flatten the load curve”, then the extent of grid upgrades is less and the costs to the customer are also reduced.  Still, it is clear that EV charging at scale will significantly impact peak load. This clearly illustrates that Demand Charges for C&I and potentially residential customers will increase significantly in the future unless we are more proactive about managing peak loads.

So, where does demand response play into this? While demand response typically addresses peak load for the utility, according to Energy Watch, it does not necessarily address a customer's peak load management problem. We need solutions that address both peak load management and demand response to minimize customer bills and reduce the investment required to transition to a clean grid.

If we manage a customer's peak loads in a coordinated way, we can simultaneously operate utility demand response programs and reduce the grid’s peak load, reducing expensive distribution system upgrades and lowering bills at the same time. However, establishing demand response and peak load management programs with a customer require joint coordination that is not always easy to achieve. Distribution grid integration costs depend significantly on how generation and load is spatially distributed, and more and more system upgrades put generation close to the loads on the distribution networks. Also, grid upgrade costs could be minimized by guiding smart load and smart generation systems into low-cost or low-impact locations and coordinating customers locally to minimize peak loads and avoid demand response events.

But how can this be done?

USING LOCAL FLEXIBILITY MARKETS TO REDUCE PEAK LOAD AND MINIMIZE GRID UPGRADE COSTS

This is precisely the complex problem that NODES is solving in Europe. Creating innovative local markets for flexible energy NODES addresses the most significant single challenge we face managing "peak load". This solution can save millions or even billions of dollars for utilities and local community customers.

Goals:

  1. Minimize the need for distribution investment to deal with increased peak load (potentially $billions in value)
  2. Minimize peak load charges for customers (up to 30% of bills)
  3. Buy time and reduce the need for grid investments

One of the biggest concerns related to peak load management deals with seasonal peaks where capacity is not available to meet demand. In hot climates like Texas and Arizona, the peak load tends to correspond with high temperature periods when air conditioning is a primary load.  Sthlmflex pilots a flexible market to address the peak load/capacity challenges in the Stockholm area in Sweden during the bitterly cold winter months. The project is led by the national transmission system operator (TSO) Svenska kraftnät (SvK) and two regional distribution system operators (DSOs) Vattenfall Eldistribution and Ellevio.

The Sthlmflex project addresses the peak load/capacity challenges in the Stockholm region, particularly on cold winter days. With electricity generation located away from Stockholm, the regional DSOs require a sufficient transmission network capacity to meet electricity demand. With a congested transmission network, the DSOs are however, not always granted the capacity they need. When temperatures are low and electricity demand increases, the DSOs require more capacity than the TSO can make available. There is in addition, a lack of network capacity at the distribution network level.

With growing electricity demand in Stockholm with over 20% plug-in EVs and the share rapidly increasing (in Norway EV penetration is greater than 50%!), the magnitude and frequency of the problem is rapidly increasing. Electricity demand is expected to grow 36% over the next ten years, resulting from a growing population, the electrification of transport, electricity-intensive data centers looking to establish in the region, and infrastructure, including the city's subway system.[1]  Think about that.  36% load increase means dramatically higher peak loads if they are not managed, putting the grid at greater risk of providing reliable energy and potentially damaging equipment.

The Swedish TSO and DSOs recognized that transmission network investments required more time than they had before meeting the increasing peak load demands. Investments needed for a transmission level to meet demand are not expected to come online until 2028.[2]

Instead of transmission upgrades, Sthlmflex was established to address the capacity challenge in 2020-2028 until the build out of the network can be completed. In the longer run, the project also aims to provide the DSOs with a market-based tool for assessing what actions should be taken to resolve congestion on the network and minimize the need for distribution level investment.

Sthlmflex meets these goals through a marketplace developed and operated by NODES, where the two DSOs buy flexibility from network users in the form of generation upturn or demand downturn. This offers an alternative to requesting additional transmission network capacity.

Notably, the two DSOs also exchange transmission network capacity with each other. When flexibility is activated in one of the DSO network regions, the DSO can use less of its transmission capacity, which can be shifted to the other DSO.

As an add-on, sthlmflex also aims to provide a route for smaller flexibility volumes to the TSO's market for manual frequency restoration reserves (mFRR).

Illustration

The DSO Ellevio predicts that electricity demand during an average winter, the TSO’s network capacity is insufficient. The problem becomes worse until new TSO network capacity comes online in 2028. The situation is even more acute when winters are exceptionally cold.

Ellevio has entered into an agreement with a local generator who can provide generation upturn if called upon. This mitigates the problem.

Ellevio owns and operates the distribution network in central Stockholm.

Source: sthlmflex presentation to flexibility service providers available on https://www.svk.se/om-oss/organisation/forskning-och-utveckling/pagaende...

The Sthlmflex marketplace

The solution includes:

  • a near term, continuous energy market (ShortFlex market);
  • functionality that allows the DSOs to register and activate long term availability agreements in the market  (LongFlex market);
  • a tendering process used by the DSOs to tender for new longterm availability agreements;
  • validation and settlement for flexibility volumes bought by the DSOs; and
  • functionality that passes flexibility that the DSOs have not used to the TSO's mFRR market.

ShortFlex market

This is  a continuous market where the DSOs purchase flexibility close to real-time, starting the day before the delivery date and up until 2 hours before the delivery period. Most purchases are focused at the day ahead stage, with additional purchases made intraday as needed.

The marketplace enables the DSOs to buy flexibility from assets located in either DSO's network region. When a DSO purchases flexibility in the other DSO's territory, the transaction's time and volume are communicated to the TSO. The TSO increases/decreases the transmission capacity held by the DSOs, respectively. As a result, the DSOs can make use of flexibility in either network region.

The Sthlmflex marketplace ' marketplace also displays the price and volume of additional transmission connection capacity when this is available. This allows the DSOs to compare flexible resources with the cost of additional transmission capacity when available and purchase the cheaper option first. When the DSOs purchase transmission capacity in the marketplace, transaction information is passed to the TSO and the TSO adjusts the capacity held.

LongFlex market

The Sthlmflex marketplace provides functionality that allows the DSOs to register long-term availability agreements on the platform. These agreements consist of both of an activation price and an availability payment. These agreements' activation price feeds through to and are activated by the DSOs in theShortFlex market. This enables the DSOs to start the cheaper, between the longer-term agreement or offers from sellers placed directly into the ShortFlex market.

Tendering process for new long-term agreements

The Sthlmflex marketplace also organizes a tendering process to enable the DSOs to enter into new long-term agreements. This includes pre-qualification of tendering parties, running the tender and drafting the contract terms.

Validation and settlement

The solution also provides a validation and settlement service for both the ShortFlex and the LongFlex market. Before financial settlement, it compares meter values to baselines[3] to assess how the flexibility that has been sold has been delivered. At the end, it invoices the DSOs on behalf of the flexible service providers, with payments reduced for non-delivery.

mFRR (manual Frequency Restoration Reserve[4])

The Sthlmflex marketplace has also developed a solution where flexibility that the DSOs do not purchase is transmitted to the TSO's mFRR market.  This service is open to flexible service providers who have prequalified with the TSO.  NODES does not handle settlement for mFRR.

Grid Congestion Locations

NODES displays "congested areas" directly in the application. These are called Grid locations, where you are allowed either directly in a map or by using coordinates to define the covered geographical area.

CONCLUSION

The electric power grid is a complex system of systems and coupled relationships between grid operations and energy markets.  It certainly will not be getting any simpler. In this article we discussed a lot of inter-related pieces of today’s electric power grid and some of those relationships may not be obvious to people who do not work in our industry.  However, using the EnergyIoT Reference Architecture as a visual, the image above provides a perspective of the different domains discussed here.

Obviously if you have been following our last several articles, we have a bit of a fascination with the concept of NODES local flexibility markets, DSOs, CCAs, and leveraging DERs for grid services and market opportunities.  We are also huge fans of electrification, but we recognize some of the serious challenges we face as we travel down this unfamiliar road and our reliability and resilience needs and expectations grow.  The idea of creating local flexibility markets to manage peak load and provide new opportunities for innovation, increased efficiency and resilience, and to enable a clean energy ecosystem is compelling.  What do you think?

 

[1] https://branschaktuellt.se/energi/25612-bristningsgransen-kapacitetsbris...

[2] https://www.svk.se/siteassets/om-oss/rapporter/2019/systemutvecklingspla...

[3] Baselines values represent expected production/consumption in the absence of flexibility sales. They are calculated prior to the flexibility transaction.

[4] mFRR is balancing energy procured by Transmission System Operators from frequency restoration reserves with manual activation (often referred to as “balancing energy”

Stuart McCafferty's picture
Thank Stuart for the Post!
Energy Central contributors share their experience and insights for the benefit of other Members (like you). Please show them your appreciation by leaving a comment, 'liking' this post, or following this Member.
More posts from this member

Discussions

Spell checking: Press the CTRL or COMMAND key then click on the underlined misspelled word.
Matt Chester's picture
Matt Chester on Mar 19, 2021

Minimize the need for distribution investment to deal with increased peak load (potentially $billions in value)

Minimize peak load charges for customers (up to 30% of bills)

Buy time and reduce the need for grid investments

What's particularly challenging is these aren't one and done goals to be checked off a list, but rather continual priorities that must be assessed and measured. Thanks for guiding the conversation into how we can best do so!

Jim Stack's picture
Jim Stack on Mar 22, 2021

I have been driving electric vehicles for over 20 years. I have also put in DAS systems at power plants as well as large battery systems. 

    Why haven't you worried about all the excess power utilities beg people to use Off Peak? With good rates and tariffs electric vehicles can make use of that excess.  Many charging site companies like Tesla and Electrify America have large battery storage at charging sites to cover the Peaks and smooth the load. The new long range vehicles are also being equipped with V2G.  They can help the GRID at Peak Times. 

    So if you have the proper rewards with good tariff prices EVeryone will respond and work together. 

Ned Ford's picture
Ned Ford on Mar 24, 2021

While I appreciate the urge to be prepared I think the eventual impact is going to be smaller than suggested here.  Most electric car owners are going to charge at home, and while there will be a strong tendency to plug in when you get home, many will be using 220 circuits, and perhaps a lot will use 110, which will delay the time of peak impact substantially.

As the total numbers of EV's on the road starts to increase, we will of course need additional services.   But as an EV owner, once I got my 220 circuit installed, I only use the Supercharger network once every three to five months.

The total amount of electricity requires to replace the entire gasoline supply is much smaller than many people expect - 12 to 20% of total current electricity.  At 220 volts, my car takes about four to five hours to charge - and I drive more than the average driver.

There might be a few hot summer nights when my charging overlaps peak hours, but for most of the year my entire charge is in off-peak hours.  It depends for many people on whether they go home directly after work, or take part in restoring the restaurant industry.

I'm not suggesting this isn't something to think about.  Just that it may not be as large an issue as some people fear.  And it may be easily addressed through a variety of simple measures.  Perhaps drivers can be given software to allow a delay in the start of charging.  One idea I think is important is using streetlights to provide public charging for people who have cars but don't have private parking.  Those would be easily controlled to delay charging, or to charge a higher price.

We're headed for an amazing few decades.

Get Published - Build a Following

The Energy Central Power Industry Network is based on one core idea - power industry professionals helping each other and advancing the industry by sharing and learning from each other.

If you have an experience or insight to share or have learned something from a conference or seminar, your peers and colleagues on Energy Central want to hear about it. It's also easy to share a link to an article you've liked or an industry resource that you think would be helpful.

                 Learn more about posting on Energy Central »