There is a test procedure for hydroelectric turbines called "index testing" that is documented in the American Society of Mechanical Engineers (ASME) Performance Test Code (PTC), PTC-18. This is accompanied by turbine control system testing procedures in ASME PTC-29. Performance limits for the turbine control systems are delineated by the Institute of Electrical and Electronics Engineers (IEEE) Std-125 & Std-1207 standards.

These procedures and standards will determine the optimum-efficiency head and wicket gate to blade relationship for a Kaplan turbine under varying head and power conditions, and assure that the control system is robust and precise in controlling the turbines by providing measurement procedures and quantified standards (or limits) for comparison. These tests and standards were developed to maximize electrical output and minimize the environmental impact of Kaplan turbines.

These are both American National Standards, approved by the American National Standards Institute (ANSI) for this type of equipment. In my opinion, these procedures and standards should certainly be adhered to by the U.S. Army Corps of Engineers (USACE) in their management of the turbines along the Snake and Columbia Rivers, but they are not.

Best practice - every single Kaplan turbine should be index-tested and optimized (or tuned-up) upon original installation, and then periodically every five years thereafter to ensure continued peak performance in these two areas. However, this is not the case in the Corp of Engineers' powerplants. Most of these turbines have never been index tested. In practice, USACE only index tests 1 or 2 turbines of each family in every powerhouse about every 7 years. Typically, operating efficiency losses of from 2% to 3% are found, resulting from misalignment of the gates and blades away from their optimum relationship.

To worsen this situation, even after the Corps index tests the turbines and identifies the 3-D cam (The term “3-D cam” refers to the feature in the turbine governor that controls the blade to head and wicket gate relationship) surface misalignment, they do not update the 3-D cam surfaces to achieve the highest possible operating efficiency for the turbines; they just leave them as-found.

Another issue is that USACE is predominantly using one supplier, Automated Control Systems Incorporated (ACSI) of Vancouver, Washington, for their control system work. There is no clear justification for the apparent favoritism towards this one contractor, which has developed and deployed the 113 Kaplan turbine blade control systems currently in use at all Corps Snake and Columbia River powerplants. The observations I made while working with the Corps on their turbines in the two worst fish-mortality powerplants, McNary and Ice Harbor indicate that the turbine control system equipment is sub-standard and the Corps is covering it up.

I know this from firsthand experience; I was there, and measured the performance of two of their turbines and control systems with my patented automatic Index Test Box (ITB) for Kaplan Turbines (Patent# 4,794,544 was awarded in 1987 while I was working for the Woodward Governor Company). After the patent expired, USACE contracted (USACE contract # W9127N-04-D-0009) with my Actuation Test Equipment Company in 2004 to buy an updated ITB from me. This new instrument was demonstrated successfully on unit 9 at McNary Dam in December 2005 and again on unit 3 at Ice Harbor Dam in February 2006.

During these tests we saw that the blade control systems that USACE’s Hydroelectric Design Center (HDC) and ACSI developed and deployed are substandard; they do not meet the ANSI/IEEE performance standards as measured by the ANSI/ASME test procedures for turbine control systems. This substandard performance manifests in unstable turbine operation - the turbine's gates and blades hunt and oscillate uncontrollably, which causes premature wear on the turbine blade trunion bearings. Replacing these bearings requires dismantling the generator and pulling the turbine out of the hole in the floor – which is a long, expensive and difficult procedure. This is especially unwelcome when the wear causes a breakdown, necessitating unscheduled maintenance on the unit.

To prevent this hunting, oscillation and subsequent wear, HDC and ACSI have added “deadband” and "deadtime" functions to the blade control algorithm. At McNary Dam, the deadband is set to 1.0 degree, out of a 15.5-degree arc of full blade rotation. By the ASME method of deadband measurement, which sums the error on both sides, the 1.0 degree deadband is actually 2% deadban ; there is also a 35 second delay (deadtime) before the 3-D cam responds to changes in head or gate. The IEEE industry standard (as measured using the ASME procedures and met by all other turbine control system suppliers) is 1.0% of the span of blade rotation, or approximately 1/15th of the deadband seen in the Corps' units at McNary Dam. The standard for deadtime is 0.2 seconds, or 1/175^th the deadtime seen at McNary Dam.

This was and is a bad solution; when people in the hydro industry learn about these improper engineering practices, there is much laughter and derision. Then a few wisecracks about "close enough for Government work," but it's not funny.

This control system deficiency, combined with the misalignment of the 3-D cam surface resulting from USACE's failure to index test and optimize each of them periodically, on an individual basis, results in lost efficiency and greatly increased turbulence in the water flowing through the turbines.

This increased turbulence makes turbine passage much more hazardous for the downstream migratory fish, which seems to be the crux of the Federal lawsuit. To describe it succinctly - during times of the year when extremely endangered species are migrating downstream, the Corps isn't allowed to run the turbines at all. Per the BiOp and your corresponding court order they must open the spill gates to minimize the hazard to these fish.

Marine biologists have agreed that minimizing the turbulence by operating at the turbine's highest efficiency point reduces the turbulence and hazard to the fish, which is a fundamental concept written into the BiOp. Thus, during periods of the year when marginally endangered species are migrating downstream, the BiOp allows operation only within the top 1% of the efficiency envelope.

The substandard performance of the control systems and the misalignment of the blades, both of which factors (I witnessed and have documented) prevent the Corps from operating the turbines within the top 1% of the efficiency envelope. Because they don't index test every turbine individually, they are not even certain where the operating efficiency peak is on each of them. The Corps just made their best guess, took their best shot, then told you they were in compliance with the BiOp terms.

This is a falsehood - by my data and analysis they're not even close.

On unit 9 at McNary Dam, misalignment of the 3-D cam surface for the gates and blades and the substandard control equipment has reduced operating efficiency by approximately 2%. They simply cannot get to within 1% of the efficiency peak of the machine. By the BiOp and your court order, they should not be operating these units at all.

The problems I'm trying to bring to light are exacerbating factors to the fish mortality problem. The industry standards are intended to prevent this sort of thing from happening, but they were circumvented by USACE designing their own control systems for these turbines and using a “captive supplier” (ACSI) to deploy this equipment, instead of purchasing it through a proper open solicitation and bidding process from reputable suppliers and specifying the applicable industry standards.

When USACE bought similar equipment from Woodward Governor Company on a legitimate Purchase Order (P.O.) in 1980, the contracting officers wrote the industry standards into the P.O. to assure they would get equipment that operated properly by specifying the industry standard 1.0% deadband and 0.2-second deadtime.

The machinery at McNary, purchased on a blanket P.O. from ACSI, has an excessive 15% deadband and 35 second deadtime.

When I brought this substandard equipment to USACE IG's attention, they excused this by telling me that because the “blanket” purchase order didn't have the industry standards written into it, there was no violation of the provisions of the acquisition documents, so there was nothing the IG could do about it. I argued the point about industry standards and non-compliance with your court order, but to no avail.

Perhaps my complaints should have been against the contracting officer for allowing such sloppy equipment to be purchased and deployed without utilizing the proper safeguards provided by the industry standards. To my mind this is a root-cause of the fish mortality problem.

I am currently reporting this situation to DOD IG, which is senior to Army IG, which is in turn senior to USACE IG. This multi-level bureaucracy has been very effective in thwarting any criticism that they do not want to deal with, but some small progress is being made.

Though not a panacea, it is a certainty that this dire situation can be improved significantly by addressing the mortality problem at the point where the killing actually occurs - at the passage through the turbines.

Sincerely yours,

______________________________

Douglas J. Albright – President

Actuation Test Equipment Company