Joe Pinkerton, CEO and Founder of Active Power, Inc.

By Farah Saeed, Program Manager, Backup Power Solutions

Joe Pinkerton is also the founder of Active Power. During his tenure, Pinkerton raised $45 million through five venture capital financings, directed the development of the world’s first commercially viable flywheel energy storage system, successfully negotiated distribution agreements with several F500 companies, and was awarded over 30 patents for innovations in power quality technologies. In 2000, Pinkerton raised an additional $138 million through Active Power’s IPO while helping to manage sales growth of 365 percent. Prior to founding Active Power he was a principle with FRC, where he directed two joint research projects with the University of Texas. Pinkerton holds a B.A. in Physics from Albion College, in association with Columbia University. Pinkerton was recognized by Ernst & Young and awarded the 2001 Entrepreneur of the Year Award for High Tech Manufacturing in the Central Texas Region. Enjoy the interview.

Farah Saeed (FS): With your current products, Active Power is operating in a very specific market segment namely the static flywheel UPS market. What really separates your company from competing battery-based UPS systems? How would you characterize end users that adopt flywheel UPS systems?

Jim Pinkerton (JP): Our focus is really on reliability and cost of battery-based UPS systems. The main failure component is the battery and the main ongoing cost is battery maintenance. Our current flywheel UPS systems, when you look at it from any kind of lifecycle cost basis, we are removing the main failure point, and thus the chief maintenance and replacement cost item, which is the battery.

The basic value proposition is lower cost and higher reliability; in addition, flywheel systems are environmentally friendly and compact in size compared to battery-based systems. But in reality, cost and reliability tend to be the main focus of end users. It is really interesting… we have shipped over 250-megawatts to 39 countries, and when you look at our customer list we have hundreds and they run the gamut. They are high-powered datacenters, military applications, broadcasting, healthcare, and heavy-duty manufacturing; so I would say we penetrate every market in which battery-based systems are used. Where we shine would be in rough environments such as India, China, or northern Africa, where there are outages every day. A transition to gensets is required, and the units could be placed in unconditioned space - that is where our flywheel systems really excel over batteries because they can just get hammered day in and day out and take the heat with no reduction in performance. But as I mentioned before, we also find applications in really every UPS market out there.

JP: North America tends to represent about half of our worldwide sales.

Absolutely. Our goal and projections say that we are growing in sales in all regions, but I would say we do have a unique opportunity for flywheel systems over batteries in developing high-growth areas such as India, China, Africa and even Turkey. These are really rough environments. Literally, you cannot use a battery. There was a customer in northern Africa who tried to use a battery-based system and within a year, it literally exploded. They explode because of the environment, which is characterized by heat and frequent power glitches. The transition to gensets is also happening. So we put our systems in and now they have thousands of discharges logged and no degradation in performance. And I think that, potentially, those are higher growth areas, going forward, for the flywheel. I will, however, contrast that with our new product, CoolAir DC, which is really hitting the mainstream North America and European markets.

JP: We are working with Fortune 500 companies in those regions. What some people do not understand is that we sell about half of our flywheel UPS systems through our partner, Caterpillar, and half under our own Active Power brand name. Most people think that the cost will be higher upfront but will save money over time. However, our own channels selling the 300 kVA system can compete head-to-head on initial price and then sell on the value proposition going forward. Most of those regions are very cost-focused so you have to be competitive in the game.

JP: I think there are a couple of differences. One common misconception is that fuel cells are a direct replacement for batteries. They really are not. They are arguably a replacement for diesel generators. They consume fuel and cannot be electrically recharged; the PEM fuel cells take about half a minute to a minute to start and what they use in the interim is lead-acid batteries, actually. So I do not think fuel cells are a true replacement for batteries. And on the ultracapacitor side, the product is price competitive with discharge times in the order of two seconds …even at fifteen seconds where our current flywheel is, it cannot be cost competitive in the high power capacity sector. Yet the CoolAir DC provides a backup runtime of 15 minutes. So it certainly does not compete with ultracapacitors. In fact, we are looking at an ultracapacitor instead of a flywheel component in our CoolAir DC. Fundamentally we can go in and take batteries and put the CoolAir in and discharge and charge electrically, with absolutely no emissions or switching out any hydrogen tanks. So there is going to be a big advantage there. At 80 kW, the pricing that we have is so dramatically lower than a fuel cell that it is roughly a factor of ten.

Jim Clishem (JC): Another component that we must consider is that a fuel cell of 10 kW power runs on basically a 50 percent cycle.

JP: That is a good point in a datacenter application, where there are some companies that are testing fuel cells. And if you generate 10 kW of back up power you also generate 10 kW of heat. And then you have water-cooling systems because you are generating tremendous heat. So I think it [CoolAir DC] has a tremendous price advantage over fuel cells and it is so much easier because you don’t have to deal with explosive hydrogen issues and it has backup runtime that an ultracapacitor cannot come close to. I also think that since we have been out there and selling systems since 1998, in the order of 250 MW in 39 countries, we have 15 million hours of customer backup runtime so customers know that we can make battery-free systems that work reliably. We have taken all of that knowledge and experience and put into the CoolAir products. So we can hit the ground running. A brand new company on the other hand has to deal with manufacturing issues such as acquiring ISO 9000 certification.

JP: No, no, of course there are. The biggest challenge is that the power quality sector is by nature a conservative one. So we are introducing a new technology in a conservative market. Our main challenge is to get over this newness factor, which in a lot of industries is good but not in this one. But I think relative to other new technologies, Active Power --is going to penetrate these markets much faster than other companies in this sector.

JP: Yes. We have our internal projections for the product. However, as a public company we have not disclosed that and do not plan to. But if I look forward a couple of years, and if this product is fundamentally offering lower total costs and higher reliability, plus provides cooling during discharge and is environmentally benign, we would have be disappointed if we did not have a double-digit share of the market by that time.

JP: I have not heard of any company that has developed anything similar. The closest thing is at the utility substation level that makes the compressed air systems we are talking about - a couple hundred megawatts. They compress an underground salt cavern at night and then use that air to feed a gas turbine engine during peak times. This process uses one third of the fuel when compared to a typical gas turbine, but that is at hundreds of megawatts. It also requires fuel and an underground cavern. But that is about the closet thing that we have seen.

JP: Our flywheel systems have a useful life of 20 years, and we expect to achieve similar performance from the CoolAir products.

JP: One internal component is a small flywheel - using about 10 percent of the energy so it offers step-loads and unloads. It also has two sets of bearings that have to be replaced every three to four years.

JP: Well, absolutely. In fact, one of the ways we are going to market with CoolAir DC is through a usage plan, where customers pay on a monthly or a yearly basis, and we will provide everything including warranty and service. There is a simple set rate and talking to a lot of customers they like that. For batteries you have a variable service charge of maybe $1,000 one year and $4,000 another year, and customers do not like those kinds of surprises.

JP: We started first with prototype parts. In fact I have in my hand the main component, which is the turbine rotor. And just as an example we started out with high-strength steel and used a very expensive process to make these tiny little blades. Since then, if you can tell, we have moved to aluminum, which is much lighter and smaller. We have also simplified the overall design, repackaging the system into an even smaller footprint.

JP: No, not really. We have actually improved response time because this aluminum component takes very little time to reach full speed. So the product is lighter, more serviceable, and more cost-effective. We engineered a lot of efficiencies along with cost reduction.

JP: We do manufacture some of the key components of the flywheel in house, but our focus is on assembly and testing. So, we outsource many of the subassembly parts, mainly from the United States. These come into our Austin plant and we assemble and test them and get them out the door. We have purposely designed the new system around off-the-shelf components, and rely on key component experts as opposed to using valuable R&D dollars to develop them ourselves..

JP: We can produce 60 to 70 units per quarter. We built our plant during the dot.com boom and invested a lot of money. This is really the premier three-phase testing plant in the world. We have 7.5 MVA of load. We can test 50 and 60 hertz simultaneously. We have the capacity from a testing standpoint to increase the flow at the factory by a factor of ten. The plant is about 130,000 sq.ft. Roughly half is committed to the flywheel and half is dedicated to CoolAir DC. We can go from just under 100 units per quarter to several hundred units per quarter rapidly.

JP: We are offering customers, as I had mentioned, a usage plan and we have designed this just under the cost of batteries at all points of time. We consider an equivalent battery to be a flooded cell or a dual-strength sealed cell. So we designed the pricing so that from day one it costs less.

JP: Absolutely. We have been talking to a lot of customers in the power sector. Between 20 and 100 kVA there are many small- to medium-sized datacenters in high-rise buildings that do not have generator backup. They are on the 29th floor, making it impractical to have a genset stuck on the side of the building. So when the power goes out the air-conditioning also goes out. So they have these highly dense 1U servers and they are going over 100º F within a minute or two. Therefore, they have to shut down these systems even though they might have an hour of battery backup, and this costs them money. So with CoolAir DC we can go in there and solve their thermal issues by expelling cool air while maintaining the power source (and I must clarify that the proposed product that can do this will be launched in 2006). So for a 30 kW load, the CoolAir DC can provide up to 30 kW backup power but it can also absorb 30 kW of heat during discharge. CoolAir DC has been sized for an hour backup and can therefore actually provide an hour of backup because it is absorbing that much heat. So that is one application.

Another application where they do have a generator for backup is a typical large datacenter. Well, with CoolAir DC in and around the server system, we can direct cooling air on hot spots of blade servers. Say you have a breakdown of a particular unit so there is only a portion of the datacenter that is getting hot, then you can make the CoolAir go into discharge and start cooling even when there is not an outage; so you can react and repair that CRAC unit and get something else online. So in essence it is providing backup cooling. In this application, you can set the temperature so that if there is a power outage on the DC bus, the CoolAir DC can be instructed to go into discharge. If it is just a particular section of the datacenter and you have access to quite a bit of stored compressed air, you could run presumably for several hours and have some reaction time.

The products that we will be shipping by the end of this quarter and next quarter will have standard cooling discharge capabilities. They will discharge approximately 700 cfm of air at around 59º F - it is a few tons of air-conditioning and it absorbs its own heat. But we have already proven that we can substantially provide more cooling - around 10 to 20 tons. That product is expected to be available later in the year.

JP: Absolutely, because right now with the flywheel we are focused primarily on the above 200 kVA market (high-power market), and the flywheel system is typically connected to a generator where it provides the energy necessary to bridge. So [with flywheel systems] we do not target every market even in the above 200 kVA and we certainly do not fit well into the bulk of the 10 to 100 kVA market that usually do not have generators. CoolAir DC immediately opens up the 10 to 100 kVA market [for Active Power] as well as the above 200 kVA market (in sectors that do not rely on generators). It dramatically expands our market. In my view, with a 10 or 20 kVA product, we can provide a compelling value proposition for virtually every market. If you look ahead a year or two, there are some other major markets we can enter with CoolAir DC, such as the cell site or telecommunications markets. Roughly there are 3 million cell sites, and they all have battery backup. The carriers are really dissatisfied with those batteries so we are exploring making a lower-end product in the 5 to 10 kVA range that can run eight to twelve hours, which is another $2 billion opportunity.

JC: To add to Joe’s comment, one of the immediate objectives is to introduce alternative energy storage such as the CoolAir DC that provides extended runtime. We also knew that by entering this broader market we needed a companion to offer a complete solution. Therefore, we sourced a GE UPS labeled as Active Power (a double-conversion UPS 10 to 500 kVA). Thus our immediate objective was to offer a complete solution.

JP: Datacenter managers mostly prefer double-conversion [UPS systems] whereas flywheels are single-conversion.

A lot of these managers want at least fifteen minutes of backup and CoolAir DC can provide that. We are moving away from niche markets and into the mainstream markets.

JC: So again, if somebody did not want to buy a battery-based UPS we do have the flywheel system as a solution.

JP: Absolutely, because as I mentioned, in the emerging regions if you are bridging to a generator a few times a day, the flywheel is an ideal solution. Especially in the above 500 kVA range it is really hard to beat in terms of power density. Our CleanSource flywheel is a lot more power dense than batteries. So I think flywheel will continue to grow in the years to come.

JP: We need to go from gross margin positive (which is something we have achieved over the past couple of years) to positive earnings, which largely involves growing the top line products in 2006 and beyond. But in terms of product development, if you look at 2006, the first order of the business plan is to be able to electrically parallel CoolAir DC. So the units can provide up to several MVA and provide that controlled cooling that I talked about earlier. Those are really the tasks in 2006.

If you look at 2007 and 2008, the cell site market is expected to demand a lower power product with longer runtime capability at 48 VDC. At this power level there are some other interesting applications such as load leveling for solar and wind applications. Solar, wind, and electric utility companies are coming to us and asking if we can launch a product that will beat batteries, which are being used for load leveling intermittent sources of power, and CoolAir DC could be a good match for that. We will explore some of that going forward. First, we would like to expand in the three-phase market, next go after the telecommunications market, and then look at some other markets.

JP: With the flywheel market we just did not want to take on the world. We wanted to position ourselves with strong companies so we teamed with companies such as Caterpillar (which is the largest maker of generators), we teamed with Powerware for our DC flywheels and sell through them, and we also teamed with GE Zenith. So we do not want to take on the giants. Ideally, we would like to partner with them. Take CoolAir DC for instance - it can work with any standard UPS. So there are so many different products that we can make with CoolAir as it is such a broad technology platform. I think there are opportunities to specialize in this technology and work with tier one UPS companies. MGE is one, and an executive publicly stated that 99 percent of the problems [with UPS systems] they see are due to the UPS battery. So if you care about your customers you want to solve that.

Source: Active Power, Inc.