EVgo & Charging Infrastructure: A Valuable Asset?

If you think that fast charging stations are high growth gas stations, you're wrong. Part 1 of this analysis digs into the unit economics of charging infrastructure.

Efficiencies is a blog situated at the intersection of strategy and sustainability, exploring all that relates to the business models and industry dynamics of the disruptors and the disruptees in our shift to the Green Economy. If you like what you read, please consider a free subscription or sharing with a friend! Join the discussion and let’s grow this community.

Perhaps you’ve noticed the SPACs¹. Nikola, QuantumScape, Lordstown Motors – 2020 was a bumper year for EV related companies looking to cash in on public markets without the drawn out process of an IPO. Trading at valuations north of 30x sales, you could be forgiven for forgetting that not too long ago businesses were bitterly resisting the EV transition. Charging infrastructure was one of the bitterest points of contention, a chicken-or-the-egg problem that kept automakers dragging their feet on EVs for years. Who should build it? Automakers should build it, activists said, to repent for all the environmental damage done by their gas cars. The government should fund this, the automakers said, we can’t be expected to build out billion dollar infrastructure for a product there isn’t a demonstrated market for. Today, charging infrastructure is proliferating and – you guessed it – there’s a SPAC for that. And this one is promising 30% IRR unit economics. So with EVs now an inevitability, is the infrastructure that powers this fast growing population a valuable asset for investors to own?

Quick Summary: EVgo

Founded in 2010, EVgo is an industry-leading owner-operator of over 800 direct current fast charging (DCFC) locations for electric vehicles in 67 major metropolitan markets across 34 states in the U.S. The company has entered into a business combination agreement with the SPAC CRIS (NYSE:CLII) with anticipated net proceeds of $575m and a pro forma implied equity value of $2.6bn. The combined company is expected to trade under the ticker EVGO after the transaction closes sometime in Q2 2021.

Revenue growth is driven by (1) footprint expansion and (2) station utilization. EVgo expects to have built over 16,000 DCFC chargers by 2027 (versus current 1,400), underpinned by partnerships with automakers, rideshare providers, government and utilities.

The Opportunity

EVs are currently 3% of U.S. car sales in January, a number expected to rise to 10% in 2025 and 30% by 2030, depending on who you ask. The projections vary, but all imply massive growth. The first adopters will be retail, who drive an average of 200 miles a week, followed by fleet, who drive hundreds of miles a day, and long-haul trucking. It’s a great set up for growth – the electrification of U.S. VIO (vehicles in operation) is a massive tailwind by itself, but growth in charging will be sustained at a higher rate by late adopters consuming multiples of kWh more than the early adopters. EVgo’s management expects a 100x growth in charging demand over the next 20 years.

So aren’t these essentially high-growth gas stations? Gas stations without the human labor? Well yes, but actually no. It’s more nuanced than that.

Levels of Charging

The first thing that’s important to understand is there are two tiers of chargers: slow charging and fast charging. How fast a vehicle can charge depends on the maximum kWh output of the charger and the maximum kWh the car’s battery can accept. There are 84,000 chargers in the U.S., of which 13,000 are fast chargers. Slow charging is categorized as level 1 and level 2; level 1 chargers are typically found in single family residentials, while level 2 chargers are what you see in commercial parking lots, parking garages and multifamily parking lots. Slow charging is, well, slow, with level 2 adding about 3-5 miles of charge per hour. This is sufficient (if unimpressive) for the majority of retail drivers who drive less than 50 miles a day and leave their cars to charge overnight. But for fleet drivers, from Uber drivers to Fedex, driving is their business; they require more frequent recharging for the greater distances they drive and they are far more sensitive to the time it takes to recharge the car. They require fast charging.

The Electrical Grid

This brings us to the second major difference. Whereas gas stations simply take shipments of oil, charging infrastructure taps into the existing electrical grid. The electric grid is an opaque system with locally set rate structures and heterogenous demands on its services that wasn’t configured with electric vehicles in mind. And that – as we will see – makes the economics of fast charging vastly different from slow charging and, yes, gas stations.

Utility rates are based on a cost-of-service philosophy, which asserts that electricity system users should pay for any costs they impose on the system. Commercial and industrial electric customers have higher electric demand than residential customers due to their heavy industrial equipment and the energy needed to power lighting and equipment for hundreds or thousands of employees. The high-power capacity needs of these customers is taxing on the electric distribution system; the system must be able to meet peak demand events, which often requires the utility to operate additional equipment to manage the electric load. Because of this additional cost, utilities levy an additional demand charge on these high usage customers. The demand charge is a fixed charge based off the highest energy usage in the monthly billing cycle².

Fast charging stations are somewhat unique in that they are characterized by high power capacity and low energy utilization. With the low penetration of EVs today, a DCFC station can sit unused for most of the day. But if that DCFC station has more than one plug it can easily hit 350kW or even 1 megawatt of electric demand if multiple vehicles charge at the same time. Because demand charges are based off the highest amount of energy used by the DCFC station, low utilization punctuated by massive spikes in power output can set up incredibly unattractive economics for DCFC stations – the lower your revenues, the higher your marginal cost. Below is an actual bill for an EVgo charger in the PG&E territory, where demand charges resulted in an effective cost to EVgo of $1.61/kWh despite the actual cost of energy being $0.178/kWh.

Below is a figure from the Great Plains Institute’s 2019 white paper “Overcoming Barriers to Expanding Fast Charging Infrastructure in the Midcontinent Region,” which shows that higher capacity fast chargers cannot break even due to the demand charges they incur.

Changes to rate systems are a work in progress and could present a potential barrier to growth. This also means that, depending on the existing rate structure and/or the proposed solution, the unit economics of fast chargers will not be uniform across the country or even across states. Solutions must occur at the state and utility level. The favored solution involves a waiver from demand charges for a period of time and time-of-use charges that encourage drivers to recharge during off-peak hours to minimize the strain on the electric grid. The implicit assumption is that when the demand charge holiday period is over, higher EV penetration will equate to higher utilization, spreading out the demand charge over more kWh and bringing down the effective price. Below, EVgo describes Southern California Edison’s EV rate structure in a response to New Jersey’s Electric Vehicle Infrastructure Ecosystem 2020 Straw Proposal (emphases mine):

“Southern California Edison (SCE) received approval from the California Public Utilities Commission in May 2018 for a suite of new commercial EV charging rates that became available in early 2019. SCE’s new commercial EV rate schedules are all-volumetric TOU rates. A key feature of the new SCE rates is a five-year holiday from all demand charges, with the expectation that EV penetration will be higher after the holiday, rendering demand charges less important. This rate is cost-based. The costs that would have been collected in demand charges are moved to the TOU volumetric rates. In years six to ten, most of the demand charges from SCE’s applicable standard commercial rates will be phased back into the EV rates, with corresponding reductions in the TOU volumetric rates. This may create longer term uncertainty for investments given the 8-10 year+ useful life of a charger, but in general is still a best practice for an EV rate.”

This rate design is structured to both allows utilities to recover the costs associated with supporting fast charging and allows the fast charger operator to earn a profit in the early years where EV adoption and utilization are low. What this means is that without permanent reform the gross profit for fast chargers are likely inflated for the mid-term in areas that have demand charge holidays.

TL;DR

What’s the point of this technical discussion? If you skipped it, here are the main takeaways:

• There’s slow charging and fast charging; fast charging is essential to widespread EV adoption, particularly from fleet and commercial drivers as well as people living in multi-family properties without access to slow-charging. Fast charging has fundamentally different economics from slow charging.

• The existing rate structures of utility companies makes fast charging uneconomical in many parts of the U.S. due to expensive demand charges; efforts to introduce EV rate structures are underway at the state and utility level. The ability to reform rates may put something of a ceiling of fast charging growth.

• Rate reforms include a ~5 year holiday on demand charges. The eventual phase-in of demand charges represents a headwind to fast charging unit economics but the extent of the impact will depend on station utilization. Rural and low utilization stations will be disproportionately impacted.

Unit Economics of Fast-Charging

If there’s one thing you take away from the previous section, it should be that due to current utility rate systems, the unit economics of fast charging stations are far more levered than people may think. But what are the actual unit economics³?

Upfront costs - equipment and development capex - are the majority of all-in expenses (70%).

DCFC charging requires specialized equipment comprised of between 2,000-5,000 individual components. As such, the component cost of DCFC charging is an order of magnitude greater than level 2 charging. Development - the other half of capex - entails preparing the site, securing necessary permits and coordinating with utilities.

Of the operating costs, it should not be surprising that 50% is electricity.

Assuming that the remaining operating costs are fixed and keeping in mind a portion of electricity costs are fixed (demand charge and customer charge - take a look at EVgo’s electric bill posted earlier in this article), we can conclude that DCFC stations have significant operating leverage. Below, EVgo presents its projected unit economics for two projects.

Not surprisingly, we see significant margin expansion between years 1-7 driven by higher utilization. It’s not indicated if demand charges are included⁴ in these two examples; the revenue per kWh declines in both cases⁵, possibly indicating that a demand charge is being spread out as utilization increases.

Nevertheless, 30-35% unlevered IRR is incredibly attractive. Whether EVgo realizes it or not depends on utilization, which is a function of number of potential EV customers (demand) and competing charging options (supply). A 22.9% utilization (assuming the station runs 24/7) implies about 5.5 hrs of charging per day. Non-Tesla mass market electric cars today take about 30-60 minutes to charge from 0-80%; assuming 30 minutes of charging per customer implies that these stations average 11 customers per day. Elsewhere in EVgo’s investor presentation, it discloses that its utilization in Denver improved to 12.91% from 5.87% in November 2019 after a deal that rolled out 100 Lyft fleet vehicles, underscoring the importance of fleet EVs and general continued EV penetration to achieve a 22.9% utilization rate.

---

This concludes part 1 of the series on charging infrastructure, covering the technical and cost side of charging infrastructure. Part 2 will take a closer look at the revenue potential and the hopefully close with an appraisal of EVgo’s realistic financial prospects and what they could be worth to investors. Stay tuned for:

• Breakdown of the addressable market’s dynamics

• The competitive dynamics of fast charging infrastructure

• Can fast charging infrastructure have a moat?

Author’s note: this article is being posted after a short hiatus. I’m still experimenting with the type of content I want for Efficiencies and what’s a realistic pace at which I can generate it. I’ve noticed that the more in-depth articles tend to be more popular than the Efficient Takes and will focus on those for the time being. I hope to settle into a 1-2 posts per week cadence. As always, feedback is welcome.

1. (Special Purpose Acquisition Companies)

2. More precisely, the demand charge is usually based on the demand measured for a billing month that is required to supply the maximum 15-minute average amount of energy used by a customer in a billing month.

3. For interested readers, EVgo published a whitepaper here that breaks down the cost economics in finer detail

4. Management didn’t address demand charges at all on the investor call. If anyone reading this has the opportunity to ask them, I’d love to know their take.

5. declines from $1.71/kWh -> $0.67/kWh and $0.48/kWh -> $0.45/kWh, respectively