Kenetech stock attracted a rare recommendation to "sell" on August 22, issued by a securities analyst with WR Lazard, Laidlaw & Mead. Share prices immediately dipped 1 3/8 to 13 3/4. The recommendation also prompted Kenetech's underwriter, Merrill Lynch, to issue two rebuttals on the following days. The rebuttals for the first time confirmed problems with blade seams, hydraulics and a few down-tower control boxes, but did not go into detail. Other problems -- cracks in blade roots and damaged generators -- have been reported by eye witnesses, whose observations are in part supported by component suppliers. The Merrill Lynch rebuttals do not mention these.
Merrill Lynch insists that problems with the 33M-VS are not serious enough to change its investment rating or its bullish outlook for Kenetech. The difficulties, including the replacement of 37 faulty blades, are described as "start-up" by Merrill Lynch. Stock rose after the rebuttals were issued to close at 15 1/8 on August 25. The year's high so far is 29 1/2 and the low 13 1/2.
The Model 33M-VS is the cutting-edge turbine under development by Kenetech since 1988 and now being marketed to utilities worldwide. It can be installed with five different outputs and three different blade configurations, according to the company. It is not clear which configurations have been installed in the hardest-hit projects, or how many configurations are being struck by failures. By September 1, the company says 300 33M-VS turbines will be operating. Kenetech reportedly spent a large sum -- $69 million -- on research, development, tooling, prototype and pre-production for the 33M-VS.
Members of the wind industry fear the consequences of any attempt to cover up the problems. They say such an attempt would hurt the industry more as utilities would judge wind as generally problematic rather than understanding the problems of one turbine model. The industry must pull together, adds one. "When is this going to catch up with them? When is this going to hit the fan? And what is the industry going to do about it?" he asks.
Investor confidence in Kenetech had been weakening before knowledge of problems with the 33M-VS became widespread. News of a negative report by a specialised investment research firm, David Tice & Associates of Dallas, had been spreading since its release on June 24. "In light of the considerable risk that Kenetech faces, we believe the stock price will soon be facing a strong head-wind," the 13-page report concludes on the basis of Kenetech financial records.
"I saw many risks, some of them general and some of them specific," says Tim Gaumer, senior equities analyst and report author. "The combination of a high value and high risk drew us to it." Gaumer stresses he is not a utilities analyst. He says power deregulation and utilities' excess capacity are industry-wide problems. But specifically, he cites items such as Kenetech's "soft" backlog of contracts that could be delayed or not even built, and its debt or highly leveraged balance sheet. He also notes its predictions of higher operating costs and more capital needed for high demand and rapid growth, already-indebted partnerships with loans from Kenetech, and the company's past performance -- including a 1993 growth largely due to a one-time sale of a co-generation plant -- and profits that have mostly been in low-margin businesses.
"They have high revenues but are not that profitable," he comments, adding that Kenetech is consuming cash at a fast rate. He also says Kenetech might be willing to sell power at a price lower than a plant owner would. "They sell plants not power so are less concerned about the cost of power than about getting contracts signed," he says.
For some time there has been a so-called "short story" out on Kenetech, which means investors are betting to profit when the stock drops, largely because of impending power deregulation, say financial sources. "There's a growing short position," says one analyst. "This has been going on since the company became public." But interest in selling short reportedly increased following the Tice report. Tice clients include more than 90 institutional investors.
Shares in Kenetech, which had an estimated $1 billion market value in mid-August, have been seen as somewhat risky ever since the company went public last September -- because of uncertainties in the independent power industry, ongoing electricity deregulation, and Kenetech's limited track record with the 33M-VS. But the company's apparent $1.7 billion backlog and impressive array of utility contracts had attracted substantial Wall Street interest, pushing the stock to almost $30 a share.
Indeed a favourable article on July 18 in Barron's -- a respected weekly financial newspaper that sells nearly 260,000 and is published by Dow Jones -- was headlined "Riding the Wind." It continues, "As implausible as it may sound, this wind power company actually makes a profitÉKenetech wind farms produce electricity at a cost comparable to juice from oil or coal powered plants." And as recently as August 22 an out-of-court settlement was reached -- for an undisclosed sum -- between Kenetech and Pacific Gas & Electric in a dispute over the utility's refusal to buy contracted power in the 1980s. In November, a jury awarded Kenetech $17.6 million in damages.
Eye witness reports
But news of Kenetech's technology problems was emerging from a variety of sources. Concerns about the Model 33M-VS were voiced when Kenetech Executive Vice President and Chief Financial Officer Maurice Miller spoke with investors on July 28, according to a transcript of the telephone conference. Responding to questions, Miller said hydraulic power units have caused "a very large part" of the low availability in Buffalo Ridge, Palm Springs and Alberta, Canada. He also mentioned the high temperatures in down tower boxes in Palm Springs. He said the problems had been largely solved and Buffalo Ridge availability was at 90-100% most days, and the machines in Palm Springs were "working terrific."
Nonetheless, eye-witnesses continued to talk of considerable downtime at the Palm Springs and Buffalo Ridge projects, with workers frequently in the field undertaking component replacements -- and several sightings of obviously damaged blades. As well as being partially confirmed by Miller's comments, these observations are supported by the Merrill Lynch reports, by the highly negative WR Lazard, Laidlaw & Mead report, and by Kenetech's component suppliers.
On July 29, the morning after the investors' conference call, Hank Hermann of WR Lazard, Laidlaw & Mead, visited Buffalo Ridge. Sixteen of the 73 turbines were not operating in a 12 to 15 mph wind, he says. Three turbines were missing blades; and blades were lying on the ground beside other turbines, with cracks visible where the blade joins the hub, according to his report, published August 22. "Admittedly, it is possible the Buffalo Ridge availability factor at the time of the visit was an aberration," he wrote. "However, while standing on the road by some of the operating wind turbines, the loud groaning, clanging, whining noises emitting from several of the machines strongly suggested to my untrained unscientific ear that meaningful problems may exist with some of these machines."
He warned of possible "significant weakness" in the stock when issuing his rare "sell" recommendation and concludes: "It is my experience, based on my understanding of the independent power industry, that it is rare for the shareholders of a company to earn a satisfactory return on an investment that encounters significant problems in the early stages of a major project or programme." Hermann, a securities analyst, currently consults to New World Power Corp, a competitor of Kenetech. He has almost ten years experience in independent power and had some knowledge of Kenetech before he signed with New World.
In its rebuttals of Hermann's report, Merrill Lynch describes the criticism as "misleading" and says critics failed to report context -- that the repairs at Buffalo Ridge were being made during a specific four day period. Visiting the Buffalo Ridge site on August 24, two days after publication of the Hermann report, Merrill Lynch researcher Don Ford reported a "well maintained wind farm operating in excess of 90%." He re-iterated the information contained in the previous day's rebuttal that 37 blades had failed, 19 at Buffalo Ridge and 18 in Palm Springs.
Yet at about 13.00 that same day, an experienced observer saw only ten of 73 machines turning on Buffalo Ridge in winds of about 25 mph. On August 25, at one point he says 11 were not operating. Also in mid-August, well after Hermann's July visit, an eye-witness says two cranes and about 15 people were working apparently full-time at Buffalo Ridge. He also states that most blades there appear damaged at the root. The root cracks are significant, he says, and from about six inches to four feet long. As a result major glass repairs have reportedly been done in the field.
Glynnis Hinschberger of Northern Power Systems, which buys power from the 25 MW Buffalo Ridge project, says she is not aware of any specific technical concerns or problems at the wind farm. "Shakedown kind of things are normal in any kind of new project," she said on August 25. "In short, everything is performing as we'd expect and we're very happy with the project."
But even before Hermann's report was published on August 22, talk throughout the wind industry was of most of the 33M-VS blades in Palm Springs being cracked or damaged, of times when most 33M-VS turbines in Palm Springs were apparently shut down when the wind reaches about 35 mph, of most of the blades on Buffalo Ridge having cracked roots, of problems with generators and hydraulics systems, and of availability lower than usual in a wind plant -- about 60%-80% compared with the normal 95-99%. For example, on August 18, out of 80, 33M-VS turbines on Kenetech's larger wind farm in Palm Springs, 28 turbines were stopped at 08.00 in winds of 18 mph, and 18 were not operating at 11.30 in winds of 15 mph, according to a long time wind expert who says he has frequently noted that kind of low availability at the plant.
Reported too was a truck load of damaged Kenetech blades, covered with a tarpaulin, with cracks at the seam between the top and bottom, in Palm Springs. And in mid-August eight turbines on one of the Palm Springs wind farms had a great deal of an oil-based substance splattered over the blades. About 30% of the blades on that wind farm were also visibly dirty.
Darryl Bone of Southern California Edison, who oversees the Palm Springs wind facilities, says of Kenetech, "They've had a few minor electrical problems. I've no idea what the availability is.... I can see they're still in start-up mode...." On August 18, he also mentioned that the winds had been "terrible" so far that month.
Blade repair costs
The first of Merrill Lynch's two rebuttals of Hermann's report was entitled "Short stories overstate problems at KWND." It said a key manufacturing process was suspended for three months earlier this year including a sanding step performed on the blade seam before bonding the unit together. According to Merrill Lynch, the 37 blades have been replaced at a cost of $20,000 -- just $540 per blade. It is not clear if this is the cost to Kenetech or the total replacement costs. "Assuming that the company produced 125 machines with three blades apiece using the faulty manufacturing process (a quarter of the machines to be built in 1994), the company's total exposure may be approx. $200,000," said the report. The cost per turbine, $1600, seems surprisingly low considering the extent of the problems confirmed by Merrill Lynch.
The $540 per blade cost compares with a price of about $18,000 for a new blade of similar size made by the British Wind Energy Group for its MS-3 turbines, according to a recently published statement by WEG's John Armstrong. A glass fibre blade from a leading manufacturer in Europe for a rotor with a 33 metre diameter, such as Kenetech's, would cost at least $20,000 to manufacture.
Kenetech's main blade supplier is TPI Inc of Warren, Rhode Island -- which made about 90% of those so far supplied to the 33M-VS. TPI's Everett Pearson says four blades have been shipped back to Rhode Island from California because of seam problems, while two had their seams repaired in the field. Of the reports of widespread root cracks on the Buffalo Ridge blades he says only "a couple" of blades have been found with a "cosmetic crack" in the faring putty between the casting and epoxy. They were re-sanded and re-potted in the field, he says. Pearson also said on August 25 that he had not heard of the 37 figure for blades replaced and pointed out that TPI does not make all blades for the 33M-VS. Heath Tecna of Washington State, which made the remaining 33M-VS blades, failed to respond to an attempt to contact the company.
Other failures
In its August 23 rebuttal, Merrill Lynch also states that about 200 hydraulic systems delivered by one supplier for the 33M-VS did not meet design specifications and are being replaced free of charge. Kenetech is likely to incur labour and downtime costs of $200,000 to $300,000, it says.
About 80% of the faulty units have been replaced. As well as the hydraulics failure, an over-heating problem in the control box at the base of the tower of some machines in desert climates has lead to a minor redesign of the cooling unit, states Merrill Lynch. No mention was made of any problems leading to generator burn out in either of the two rebuttals.
But problems leading to malfunction of some generators are confirmed by the supplier, Marathon Electric of Wassau, Wisconsin. "The problems are not in the generator, they're in the total system," says senior product engineer John Watson. "The windings are shorting out." He also denies the problems are catastrophic. "It's like anything else; things do fail," he says.
Responding to a series of detailed questions, Kenetech's Bud Grebey says: "It doesn't warrant a response, because at this point it is suspicion and a level of tactics from competitors that we would rather not respond to. I don't want to get into an ethics battle." Another Kenetech source was just as adamant. "They're important and significant [problems] but not earth-shattering and not throughout the fleet," he said in mid-August. "Kenetech will not be overwhelmed."
A member of the financial community, without ties to any wind company, comments on the WR Lazard, Laidlaw & Mead report: "I think they've raised some fairly severe questions here." He adds: "There are rather important technical issues that are potential problems. In view of the money [Kenetech has] spent on it, it seems to me they have some pretty serious problems." He says Kenetech has a poor record of profitability, and that such operational performance could be significant. "You'd think it would be working by now, but it doesn't seem to be," he says of the 33M-VS.
One long-time industry member says Kenetech's problems do not appear to be start-up. "They've done just about every major component and I don't see that as a teething problem. You might have technical problems, glitches or control problems, but you don't have blades coming apart at the seams or losing whole generators." Another source says, "If you spend $69 million on developing this cosmic technology, you shouldn't have these problems." He adds, "It's unfortunate. It just lends strength to our opponents' arguments."
Randy Swisher of the American Wind Energy Association comments that he recently visited Kenetech's newest wind farm, owned by Sacramento Municipal Utility District (SMUD) in northern California. "The hardest evidence yet I have of the 33M-VS is from SMUD. They were running with availability in the nineties and a 67% capacity factor in July," he recalls. Swisher says he has not seen enough data to know if problems are normal "shake-out" when a new technology enters the market or if they indicate a deeper design problem that would be costly to correct. He says of course the Model 33M-VS is important for the industry. "We obviously hope that it does perform well and would be concerned with any problems with its operation," he concludes.
.png)
.png)
