The dark blue text is the Contracting Officer’s Representative (COR)
The light green text is my words back to him.
The main point is that I was separating work on the ITB source code that is protected by the Copyright from the work I was doing for HDC under the government contract.
At the end when HDC lawyers were moving to sieze the software becuase of false claims that I wrote the software source code at government expense, the U.S. Copyright proved the software pre-dated the government contract and therefore was written at private expens and not covered by the contract.
Doug Albright 2011-12-05
That’s why I wanted to do this last Mod (P00002) as Firm, fixed price. That way no matter how many hours you spend on it you only get $xx.
Dave – That’s OK with me, I don’t mind working on my product on my time after the sale to make it fit my customer’s needs better. Giving a latest-and-greatest version to your guys is good for me – their feedback is very helpful.
Even as it stands now, you don’t get paid for any extra hours either, because they were not authorized by me first. You’d have had to e-mail me and say: Dave, I’m out of $ as of xxxx date. If you wish me to continue working on this, I estimate it will take xx more hours, or $xx more dollars.
I prefer to work on my software on my own time to avoid conflicts about whose software it is when it’s done. If I charged you for any software work, you’d own the software. I like it better this way.
Then I’d have said, we ain’ got n-o-o-o-o-o money, honey! Then you: Ha, ha, ha! You k-n-n-o-o-w-w what I like! Chantilly lace and a pretty face, a pony tail....
Tom Murphy already told me they’d he’d fund whatever you guys and me agree to do. There’s money, Dave.
You know how it works. Our funding doesn’t meet your needs. Your needs have to meet our funding. One way street on that one. Always will be with Uncle.
The customer is always right. So what’s your point? That’s the way I work with all of my customers. You’re not saying anything new here.
For example, Rod wanted me to change the design from using my ATE-150 box as the signal conditioner to using a 4-20 mA transducer instead, which is not covered under any of our written agreements. No Problem, I explained how to do it to Dan Ramirez yesterday.
Dan called to ask how to connect the ITB Winter Kennedy differential transducer input in series with the 4-20mA loop transducer that he’s using for his equipment.
I drew up a schematic suggesting the way I’d set it up, detailing the dropping resistors and grounding scheme for adding the ITB to his normal equipment setup, then left it up to him to decide what to do. After thinking it over he decided to parallel the ITB to his system at the transducer plumbing, instead of behind the transducer in the wiring to keep it more simple and goof-proof. That electricity can be real tricky.
Rod tentatively agreed that if I did this for you for free, I’d get a copy of the data from the next few upcoming tests, some Kaplans and some Francis turbines. From discussion in our last meeting, it was also suggested I should receive this additional data as replacement for the unacceptably truncated data set from the September Field Test.
Your equipment’s limitations denied me the 2-degrees of rotation that was planned for the index test, and we were unable to move the blades far enough in the flat direction to see the efficiency crest over. I’m preparing a magazine article using this data, so a full set of data for “show-n-tell” would be a big help. It just wouldn’t do for me to need to explain how and why the data was truncated in a fancy magazine article.
We won’t put any software we do not have all the code for into any SCADA system. To do so would be... irresponsible. So it’s source code or we develop our own indexing software, right into the SCADA system. And since they can probably develop their own for less than $750K, maybe even much less, why would they want the ITB? Also, even with source code, someone would still have to go through it line by line to assure no malicious code (SCADA sys). Which makes ATEC even more expensive.
If you think they can develop their own for less than $750k, then why haven’t they done so already? Many of your people have told me you’ve tried and failed at this before several times over the last 20 years, including Tom Murphy when we met at the Project.
Seeing me do it once and having a working prototype will certainly be an aid to whoever is working on it, but it is dubious that your people will be able to pull it off; as I understand it, they’re already busy working on T2 optimizer system.
And then there are the other technical problems in your existing equipment.
To properly enjoy the benefits of index testing, it is required for that the blades to be positioned accurately and quickly in response to changes in head and gate. The equipment I saw at your facility does not perform so. The combination of the below listed characteristics causes the turbine operating point on the gate-blade graph to spiral in on the operating point, instead of tracking the desired position robustly.
The control system performance documented during our field test is sub-standard. Intrinsic problems with the control scheme on the 3-D cams preclude accurate, robust & responsive unit operation.
Your people say they use a 1-degree Deadband to prevent any blade motion and resulting wear on the blade mechanism. This is misleading. The ITB display indicated that the blades never sit perfectly still while running “on the governor.” There’s a normal rocking motion in some governor systems in response to external perturbations and we could see it on the ITB display.
In the real world deadband is a “peak to peak” measurement, so what your people are calling 1 degree, is actually 2 degrees of deadband, out of a total blade rotation of 15 1/2 degrees; that’s 13 percent added, designed-in hysteresis.
Total system hysteresis becomes the sum of this added deadband and the normal, unavoidable system hysteresis and slop in the blade mechanism.
By design, your system has an extra 1-degree error built into it, that’s 6 1/2 percent. This is much worse than nominal industry performance standard, and could unabashedly be called “Sub Standard” when compared to other commercially available blade control systems that meet recommended IEEE 1% blade error guideline, as measured using ASME methods and procedures as metrics.
Another red flag is the 30 second lag time seen before your blade controller output stepper motor moved in response to gate inputs. A proper blade control system will respond within 10 milliseconds, always keeping the blades where you want them to properly align the equipment to the water flow vectors in the unit. This will reduce turbulence, lost power and dead fish, or it would seem.
The ITB documented a 3-D cam hang-up on 8-September 2005. When I asked to run the ITB overnight again during our December Field test, Rod told me that overnight running had been preemptively denied. He wouldn’t explain any further. Now why would anyone not want to see unattended, overnight testing of an automatic Index Test Box for Kaplan turbines, when this was in the contract as criteria that the ITB was to be judged by?
I’m going out on a limb here, and reading into isolated facts and things I’ve heard to present my best guess as to what’s going on - I’ve heard that after the December field test, Ed there wanted a “cure letter” written against me because the software crashing problems I had violated the contract. How could he be critical of the ITB not being able to run unattended, when they had preemptively forbid an overnight demonstration of it doing so?
We saw that sometimes there was significant difference between the 3-D cam software’s “Ideal Blade” angle and the 3-D Cam data table graph. In order to learn where they were diverging, I asked to see the source code for the 3-D Cam so I could figure out why there’s a difference. Your lawyer told me that in the interest of “anti-terrorism” and national security, I could not see this critical infrastructure documentation.
The paradigm shift in your people was noticeable when they realize that the blades are actually always in constant motion, in response to the normal controlling action of the gates under control of the governor, not sitting still as previously believed.
They were mislead by the non-moving stepper motor and how slow and small these motions are in the governor cabinet; that they perceived that the blades also weren’t moving. In fact, they’re always moving (albeit a smaller amount) in response to the shallower base-line rise on the split-cam mounted on the constantly moving gate restoring shaft. It takes an instrument like the ITB, an oscilloscope or strip chart recorder to see it.
It was suggested that perhaps the real and unavoidable blade hysteresis is large enough to consume the motion we were seeing in the blade restoring system before it actually moves the blades. Maybe.
The common belief is that the halted stepping motor prevents all blade motion, and thereby reduced wear on the blade mechanism is false. The truth is the lack of stepper motor input can only reduce total blade motion; it cannot actually stop all blade motion and wear.
They had overlooked the normal, lesser motion on the blades imparted by the rise on the split-cam. While the governor regulates Unit output by moving the gates, it is also rotating the split cam on the gate-restoring shaft. The blades stylus is riding up and down on the split-cam rise and is still moving the blades, but a smaller amount because the stepping motor is not adding Its part to the motion yet.
That was redundant in some parts. Sorry.
I’ve heard about three failed Digital Governors you and my competitor tried to build at one of your projects.
I know that ACSI’s Dan Perrier tried to buy a working governor design from American Governor Company, and that Dick Hollenbeck of SoftPLC Company also tried to buy a working governor design from L & S Electric and North American Hydro. All three Wisconsin firms declined.
So you are saying you want to let them go 3-for-3? A substandard 3-D cam, nonfunctional Digital Governors, now they get a shot at an ITB that doesn’t work too?
In truth, I don’t believe DOE has the time and money for an additional ITB development cycle for whatever they can build for you, while they have the knowledge that my working, proven system is ready to deploy now. You’re bluffing.
Finally, ATEC is the only company capable of maintaining this system as it stands. A risky way of doing business, and not one HDC would care to pursue.
They don’t want to pay a license every time, or have someone else control access.
All valid points, I think. Would you care to revise ATEC’s offer as stated below? Yes, let me make a few changes.
I still believe that the $750k and geographical restrictions to give up the source code is a fair price, based on the time and energy I have invested in the methodology and software, and a bargain compared to the increased revenue you will get from your turbines.
I also believe that the additional $250k is a fair price to make it work in C++ as a GDACS module. It was an over sight not to include the source code above in this offer. Combine the two and there’s your price schedule for buying the source code for GDACS.
Please note, I didn’t offer to sell the source code, you asked for a price. I would rather not sell the source code outright. Instead, I’d prefer to sell you the complete ITBs (per the pricing in my last Email on the subject) ready to use, and the executable programs per the schedule in our contract.
Back to that GDACS SoftPLC coding suggestion. You’re not keeping your eye on the prize here.
The goal is to index test the turbines and change the cam surfaces on the turbines as expediently as possible.
With the current configuration of the ITB, it is possible to start doing this tomorrow – oh, guess what, Dan Ramirez is out doing just that, right now through the end of the week. He gave me the comfort yesterday that if everything doesn’t work right, he’ll go back next week again to finish up.
It is not necessary for the ITB program to reside inside the SoftPLC as a rung on the relay logic ladder to do so; I reject this notion of “compatibility” altogether as Empire Building, and will want to take it to a higher authority if it prevails. It’s just a SCADA system.
The complete unit price has a few man-days in it to make modifications and improvements as we go along, so I’m very comfortable with the contract we already have. The source code price was something you wanted, not something I asked for.
If you buy the source code, the additional amount for SoftPLC deployment will include all documentation and assistance necessary to get the system up and working in your GDACS network.
In any agreement, I’d also like some assistance getting work from usace Kaplan facilities in my area. Everyone I’ve contacted around here indicates HDC must initiate any work. I called Clarence Cannon and DOE about fixing their Kaplan turbine, but no - they’ll spend $2.5M without blinking to fix the crane in the turbine bay, but won’t even lift the phone to increase turbine efficiency by calling HDC and requesting an index test.
The problem is obvious, we can fix this.
David Ebner, A/E Contracts
Army Corps of Engineers, Portland, OR
(503) 808-4611, fax-4605
"The Best Information Leads to Success"
From: DudleyDevices@aol.com [mailto: DudleyDevices@aol.com]
Sent: Tuesday, January 17, 2006 12:41 PM
To: Ebner, David A NWP Subject: Re: RE-Invoice
I'll work it any way you need. There is a rub, however.
From the 22nd of November to the 3rd day prior to leaving on 11 December there was only 3 days worth of money in the mod. How can I account foe and/or bill for that period when there wasn't any money to pay for any work done? Am I to account for time I worked that you didn't pay for?
There was enough money in the mod to make the trip, 3 days before to prepare and 4 days after I returned for reports. Other than that I was working on my own nickel.
If you are looking to account for delays in getting the ITB project wrapped up, I was working on the software from the 22nd of November to the 9th of December, just not under the control of your contract. During this time I completed the transducer buggy and shipped it to McNary, again without pay. (see above - no money available) The records will show that I did respond immediately to any emails or phone calls regarding the project to prevent any delays, but again, on my time. I've been working on the PostProcessor all day, every day since returning.
I was working on the software anyway, so not charging for it keeps the struggle over source code from getting in the way. To "calendar it" - I worked every day, all day pulling this thing together.
I spent the 3 paid-for days packing and rehearsing before going. After I got back, the available paid-for time was spent on the step-by-step and the attached data report.
While I'm at it, let me broach the subject of the $750k figure you asked after getting it in writing.
Our original discussions at that price were for all Kaplan turbines in Snake and Columbia River facilities and 200 units for the rest of USACE, but as I recall, we were haggling about the PUDs and whether or not they would be included in that price when we switched to the time and material model we're working under now.
That broad brush conversion planned the original $160k to configure for GDACS, demonstrate and deliver one ITB, then the remainder was amortized over software for
320 units; 120 for Snake and Columbia units and 200 for the rest of USACE. We didn't budget for the daily pissin' matches I had with Ed over everything.
As I understand your request on Friday, the number you're looking for now is for all turbines under USACE, all PUDs and probably all Bureau of Reclamation units as well. That's a much bigger target field than the $750 was intended for. It is rather grandiose of me to think potential applications would go that far, but the possibility does exit.
My guess is, using the business model HDC has always used, whatever is purchased from me will be going directly to ACSI, who, along with Ed and others have demonstrated they are not to be trusted with anyone else's intellectual property. I would not expect them to honor any contractual restriction against full commercialization of the ITB software, were they to get their hands on it. Enforcement of any contract would be on me, so I'm uncomfortable doing business with people I don't trust.
So let me make a counter offer...
I can write C-Code to have compiled to an executable that will run in the GDACS as another rung in the Relay Ladder Logic code. If I can connect the ITB to GDACS, why can't I convert the ITB code to a GDACS module myself? It would be much more expedient for you to have me write this code than it would be to have someone else learn and understand the ITB functions and make it work in some other platform.
So instead of $750k for a dubious proposal to have me turn over my source code to people I don't trust, let me suggest a price for me to write a GDACS compatible module at about $250k. I would still use the integral password to control distribution, but I would allow you to install it in every GDACS computer you have in the field now for $1 each.
Attached are the Final Field Test Data Report, my response to your List of Items to Do and Greg's credit card statement and plane ticket.