Sunday, February 17, 2008

Future of Transportation - PART II

n the second part of his Future of Transportation series, Shai Agassi discusses the "Technical and Financial Implications" of the model he has suggested in addressing the transportation needs in the 21st century.

Technology and the Financial Implications

Apart from crossing peak oil, another major event happened around 2005 - the emergence of a new generation of batteries - Lithium Iron Phosphate (LiFePO4) - able to sustain more charge cycles and based on safe chemistry that can be put into a car. For the first time the total cost of energy for electric transportation has crossed under the cost of fuel when calculated on a per kilometer basis. The fundamental technology and economic drivers behind these two events will continue to drive the price vectors for fuel and electricity further apart in favor of the electron and battery. Within a decade, the cost of energy for a single year of fuel supply for a combustion car should cost more than the cost of energy for an electric car’s entire life, even when taking the cost of battery into consideration. The “cross-under point” had gone almost unnoticed in the world of automotive design which was focused on the hybrid-car race, yet its effect will change the industry in the most disruptive economic shift ever experienced in history. Cars are not complete products, as they would not provide any function without fuel and variety of services (such as maintenance). As the price of crude oil increased, it drove the price of fuel at the pump higher to become a much larger component of the total cost of car ownership. To illustrate, an average European car costs 12,000 euros to acquire, yet over its 12 years of life will require approximately 30,000 litres of fuel costing roughly 35,000 euros (assuming fuel prices do not continue to increase even further). In other words, we now have a container for energy built into the car - the fuel tank - costing US$ 100 to build; yet our energy costs three times the price of the car.

Contrast that with the electric vehicle where the container for energy, in this case a battery, costs roughly 7,000 euros, yet the electricity to run the car costs 2,000 euros for the entire life of the car. In the aggregate, energy to drive an electric vehicle now crossed under 10,000 euros. Historic trend lines for battery over the last 25 years shows a 50% price per kWh improvement every five years, stemming from technological and process improvements. We have seen similar effects in the chip industry, where Moore’s law predicted chip improvements amounting to 50% reduction every 18 months. Similarly, we see the price of renewable generation declining over the years, to the point where large solar installations cost today 2 euros per Watt, shedding price roughly at the same rate of 50% every five years. Projecting forward to 2015, we should see the cost of the battery and solar generation sufficient for a car reaching combined cost of 5,000 euros. By the end of the decade that price should drop to 3,000 euros with the battery and solar generation both outlasting the car. At some point during the next ten years, the total cost of electric energy (with battery) for a car will equate the cost of fuel for a single year. We predict that at some point in time before that next cross-under point the entire car industry will tip to electric drive as the main design principle for new cars. What is missing for this transformation to happen today? Infrastructure and scale. Consumers simply cannot buy products that are not available. Electric cars as well as all their critical components are produced in small runs, not on commercial scale.

The car industry was caught by surprise with the sharp increase in oil prices as the US makers focused on ever larger SUVs and vans. Even Toyota was surprised with the success of its hybrid Prius line. While the entire industry scrambled to catch up to the hybrid wave, every one of the car majors assuming no new infrastructure will be in place for a pure EV decided not to produce an EV until the emergence of a battery that can last for 10 years and provide enough energy to safely drive a car for 500+ kilometers. Since such a battery is not in existence, ( it is most likely that it will not be there for another 15-20 years) all makers pushed their EV plans into niche solutions focused on fleets of cars that run predetermined routes and come back to home base after 100-150 km, such as postal delivery trucks. It turns out that the solution does not stem from a more powerful battery.

Rather we propose the creation of a ubiquitous infrastructure that can enable a car to automatically charge up its battery when parked, and on the exceptional long drive using an exchange station where an empty battery is replaced with a full on in automated lanes resembling car-wash devices positioned in gas stations across the country. We for the first time look at the car battery as part of the infrastructure system, not part of the car, much like the SIM card inside a cell phone is part of the network infrastructure which is residing inside the phone. Since the car owners do not own the battery they can freely exchange it as needed, not fearing the issue of receiving an “older battery” in exchange for a new one. The collection of park and charge spots across a country or city, together with software that controls the timing for charging the cars, creates a smart grid - synchronized and extending the country’s existing electric grid, matching excess electricity on the grid with the need to charge batteries flattening the demand curve in the process. When we put together the charge points, the batteries, exchange stations, and the software that controls timing and routing we get a new class of infrastructure - the Electric Recharge Grid (ERG). A new category of companies will emerge in the next few years which will install, operate and service customers across this grid - called Electric Recharge Grid Operators (ERGOs).

The business model for such operators will be similar to that of wireless phone operators, and so we can predict that a few years after the ERGOs, we will also see the emergence of virtual operators on top of the physical grid (or VGOs). The economics of large infrastructure operators call for massive investments up front, which can be monetized over years through subscription-based services to consumers. Similarly in this case, once a grid is installed to the degree of sufficient ubiquity in a contained region, car owners will be able to subscribe to a complete commute solution – car, energy and maintenance contained in a singe predictable monthly price. Not only is the price predictable (unlike the case of fluctuating oil prices), depending on the length of the subscription the ERGO can subsidize the cost of acquisition of the car. As the costs of battery and clean electricity will continue to decline over the next ten years, we can easily foresee enough subsidies in the contract to the point where electric vehicles will be given for free to long-term subscribers. Assuming that subscribers will be happy to pay the same amount they pay for fuel and maintenance today, the economics require a contract lasting six years to get a free SUV.

By 2015 that same monthly fee will require a contract lasting four years and in 2020 that contract will already be reduced to three years – the average leasing contract today. Such radical process has happened before in the wireless phone industry, where it is almost expected today that a basic handset will be handed for free with any new subscription. With infrastructure and economics in place the demand curve for such new transportation model will grow exponentially – taking a significant portion of the current global demand for cars, standing at 70 million new cars a year. The supply curve for components and cars will need to scale similarly - scaling an entire set of industries, from batteries to motors and power electronics. To illustrate the rate of growth, today’s Lithium based battery market (working mainly for laptops and cell phones) produces enough batteries to power roughly 100,000 electric vehicles. Reaching a market of 10 million cars (representing only 15% market share) would require a 100-fold increase of annual production capacity globally.

On the other hand, as electric drive trains have less moving parts than the combustion engine and its supporting components, the markets for today’s mechanical auto parts (such as spark plugs and carburetors) will start to decline sharply, as will the market for car maintenance. We will also predict that used car prices will at some point decline sharply, as a result of the availability of cheaper all new clean cars, costing consumers less than the cost of fuel based used-cars. The magnitude of this disruption is discussed towards the end of this document, it will take time, but we believe it is almost unavoidable at this stage. Since we ran out of cheap oil and all new discoveries are in deep oceans or troubled locations in the world, we have now a floor price to the production and refining of oil. That floor price together with climate-related tax policy will make sure that even as demand for fuel subsides in following years as the predicted events unfold, the price of fuel will not be able to go back below the 1 euro per liter at the European pump.

No comments: