Airdrie , AB, Canada
This article is written by Harvistor LLC, Starwind5 with Copyright 2015 All rights Reserved, Worldwide. Author: Robert Reive.
UGE, Going where many small wind turbines have gone before?
UGE, Urban Green Energy is a small Wind turbine vendor based in New York City, USA which has recently received coverage in the media for it’s installation of two wind turbines on the Eiffel tower in Paris, France generating lots of hype and attention for the company. Of course such hype always is bound to come under the lens and scrutiny of the experts both real and imagined, and it has, this time by one of the Wind Industries most respected experts.
Mr. Paul Gipe of www.wind-works.org , is a renowned wind industry expert with a genuine pedigree of experience and expertise which goes back to the 1970s in US Wind and more lately a renewable energy expert called on to help build various FIT “Feed In Tariff” programs for various provincial, state and country governments is indeed one of those critics if not the best critic to shed some real light on the reality of UGE’s performance and certification claims versus the published hype.
In our opinion, Mr.Gipe’s gripe with UGE specifically around their performance and certifications claims is totally with merit, as well his facts are correct in comparing UGE versus the market leader Bergey with regards to the amount of turbine build material and labor used versus the sweep area and the power produced. Clearly in this UGE comparison Bergey wins easily.
Disclaimer: We (Harvistor LLC the inventors of Starwind5 previously known as Darwind5) do subscribe to Mr. Gipe’s newsletter and visit his http://www.wind-works.org site for research and education purposes. In our opinion it’s a great source of factual an anecdotal information collected from real world projects and the people deploying these wind projects, and has been helpful in focusing our company on what is important in building an effective small wind turbine regardless of the technology or design.
Let’s now look at a list of the issues Mr. Gipe brings up first, re: the Eiffel tower mounting of UGE VisionAir VAWTs which look like they are about 4kW in nameplate power rating:
Siting: “Inside the Tower” (N.B in our opinion this is not a workable siting to produce any meaningful power)
Purpose: UGE Art vs. real useful power production (no comment as we have no data…)
Material Used: UGE vs Bergey, about 4X more (Bergey Excel 6kW Nameplate)
Sweep Area: vs. Bergey about 50% less
Brake Control: UGE has No means of passively braking the system in hi wind vs. Bergey furling out of the wind
Rated Power output @ 11/ms : Bergey delivers 22% of Nameplate vs UGE 11% of Nameplate (4kW NP used?)
Certification: UGE When? still Work in Progress? for 3+ years, hmmm
This is all great commentary and comparative fact from Mr. Gipe which has helped us build Starwind5 to be a better product than the first version of our system Darwind5.
(The first version of Harvistor’s VAWT what we now call Starwind5, was operated as a field prototype (1.2m horizontal diameter and 1.9m height) to validate our in house configured (using QFLR5 airfoil design tool making use of MIT(Mark Drehla’s xfoil code) design. The performance simulation code was first developed with Openfoam (UK Imperial College) using the Openfoam plugin RANS solver, operating on supercomputer time lent to us on trail by Sabalcore of Florida, USA to discover lift/drag and turbulence of the Darwind5 first visualized via post processing using the Paraview(Sandia National Labs) Post Processing Visualization tool or CFD (Computational Fluid Dynamics). To create the 3D mesh necessary for modeling our system in both the RANS Solver and the Viewer for the Darwind5 telemetry and geometry we used the services of University of Waterloo’s CFD & Aeronautics expert Dr. Amir Baserinia, together with the services of Michael Hudson, Hon. Sc.- Mechanical Engineering from Queens University for the mechanical design, which was then implemented in 3D CAD and 3D Mesh models by our Senior CAD Technologist at the time, Chad Lambert. The actual design elements and concepts were my own (with some key influences, financial, technical as well as moral support from my parents together with a few creative shoves of encouragement both). Ms. Valerie Hopkins, a key player, investor and a director of Harvistor LLC, also was instrumental in capturing all the thinking and work on Darwind5 and also Starwind5 , having video logged, organized and documented the Darwind5 design and field test data generated by the tech team from several simulations and tests. Sincve that time we have moved the entire design effort on to QBlade from TU Berlin supported with FreeCad driven output to both Autodesk and Solidworks file formats to aid manufacturing efforts)
In those days we learned much from the Darwind5 applied research, design, competitive analysis, supercomputer driven simulations, real world critic technology comparisons and field test results. Suffice it to say, we had an aeronautical design performance break through moment in February of 2012 in the Darwind5 design which manifested itself accurately in the field tests validating our design and simulation logic, which is now part of the overall Harvistor LLC Copyright 2015 All Rights reserved Claim we enforce in our partner agreements and publicly for Starwind5. Shortly thereafter, the hard decisions and hard work to commercialize the applied research break through began what has become a 3 years+ journey as of the writing of this blog. As it takes much money to create something new and we did not have much of that, a decision was made to move back to the original licensing business model. 3+ Years later we have 4 partners on the verge of releasing their own products and energy solutions which utilize Starwind5 as auxiliary charging and direct grid connect power generation elements of the their own hybrid solar, wind, battery and grid connected solutions. It’s safe to say the old VC adage about startups applied at Harvistor, 4 years to get the idea formed into a workable state, and another 4 years to get to meaningful revenue generating physical products and related services…for the most part seeded with Angel/friends of money, jump started with a Canadian National Research Council grant on the premise we would fix the Darrieus design with supercomputing, with buckets of sweat equity thrown in to get us where we are today, ready to produce a few production examples for the early orders we now have. So that is Starwind5 as of today, now let’s get back to UGE, Bergey , Mr Gipe and the legitimate gripes and how Starwind5 compares. 🙂
We should also mention for reader inspiration and to recap the lessons learned we do indeed stand on the shoulders of those who generated the simulation and field testing data for the earlier Darrieus Eggbeater designs before us, MNR, NRC, NREL, Sandia, NASA, USA DoE, USDA (in some form from 1966 to today) and of course Flowind(1980-1997) and George Darrieus (1925) himself (as well as Savonius (1929)). We learned very much from this “old data” and field experience scraped up from early sources of such information hitting the web since about 2004, without which we would not have progressed to this point.
The journey from Darwind5 applied research to Starwind5 commercialization on our own nickel has not been easy. Much has been sacrificed and our early investors have been largely patient and a few have been hugely supportive for the past five years. We continue to learn a lot from the market, it’s critics, and competitors, and we have incorporated what we have learned over the past three years (it’s a journey) into the improved Starwind5 as follows (which we think address Mr. Gipe’s concerns.) :
Siting: “Mount Inside Semi Laminar Flows at Lower Heights” producing Bergey power levels at 25% lower heights
To use the standard 80 ft tower which really sees Bergey perform at it’s advertised best, we know with Starwind5 we can produce the same power at 60 ft., This means 35% less tower and foundation costs.
Purpose: Yes we are better looking functional Art than either Bergey or UGE and have unmatched real useful power production, at least everyone who has seen the design tells us that, so allows a small moment of indulgence, 😉
Of course this Starwind5 performance power claim needs to be validated by a ‘third party, and we are quietly confident this will happen this coming year self financed with real orders from happy follow on customers and integrators, the looks department is for you to judge.
Material Used: A HAWT will win this comparison you would think, having said that Starwind5 at 6kW Nameplate the 2842 will weigh only 580 lbs, this is about the same weight as the Bergey Excel 6, so go figure, UGE is the under performing super heavy weight in this category trying to keep everything nice and rigid and needing a lot of pole and foundation to do so. I Guess an H Frame rotor design that is both Darrieus (UGE) and Savonius (Windside OY of Finland) is pretty material efficient after all. Still Bergey has volume production and a 10-15% “turbine only”, cost advantage until Starwind5 production achieves the same volumes in manufacturing. (that will take a few years as Bergey is the current small wind leader by leaps and bounds)
Sweep Area: Bergey advertises a Sweep area for the Excel 6 of 16 m2, the reality is only 65% that is utilized as the center portion is moving too slowly during rotation to contribute any usable lift so to be more accurate Bergey is really more like 10.7 m2. Now how do you measure the sweep area of a symmetrical elipitical 3D shape ( symmetrical egg?) which is what Starwind5 most certainly uses to create sweep area? Certainly no 2D measurement system would work. The reason is lift occurs continuously as a system on both the leeward and windward surface of Starwind5’s rotor system AND in low and medium wind speeds 2.5 to 5m/s the Optiflow Savonius cups also create positive push rotation to ensure consistent startup and gust capture where TSR “Tip Speed Ratio” is less than 1 (that is the rotor blade is moving slower than the wind speed), Starwind5 2842 is some respects is measured the same way as Bergey, adjusted to remove 35%, 17.5% from the top and bottom to measure only 65% of the effective power band generating lift and positive push which wraps around the working equator of the Starwind5 rotor system. Given the real diameter of the 2842 is 3m and the rotor blades are in a state of lift and positive push at least 65% of the rotation on average, by using py to calculate the circumference of the lift and push range for 65% of the height of the Starwind5 Model 2842 we can figure out the effective “cylindrical like” sweep area. To be safe we use only 2.9m diameter and conservatively calculate the sweep as 6m (allowing for effective sweep area shrinkage due to the blade curve) multiplied 65% of the height which is conservatively pegged at 2.7m which equals an effective working sweep area of 16.2 m2. Hmm, this means the Starwind5 Model 2842 has 45% more sweep area in a visually more compact area , where the swept blades generating only 15 db of sound 15 feet from the source. Let me see , 10.7 m2 is about 60% of 16.2 erring in favor of Bergey, which means Starwind5 has 45% more usable sweep area? Yes, this is the “ah ha” moment, and why Starwind5 will out produce Bergey at a 25% lower height by about 35% more power.
Brake Control: Starwind5 uses increasing solidity at higher speeds together with fairing deformation of Optiflow cups to increase resisting shedding eddies to provide a very effective, trusted, 100% reliable means of passively braking the Starwind5 system in hi wind and keep producing power at hi levels up to 24 m/s vs. Bergey which is furling out of the wind at 15 m/s and dropping power production in high wind by over 65% at a minimum in these high winds which contain the most power. Starwind5 continues to operate at rated peak performance of 6kW up to 24 m/s before Starwind5’s self regulating design stop increasing rpm at this level. Interestingly as the Starwind5’s Egg Shape becomes more solid with solidity increasing with speed from the top and the bottom towards the working equator and the rpm levels out at 24 m/s+ the load on the tower also is much less than UGE because of the visible shape as the cross flow wraps around these increasing points of solidity on the outside of the turbine at these points and not through it, which means tower bulk and foundation type is much smaller and less in cost by at least 50% compared to UGE, and at least 3% less than Bergey given the lower mount height and less overall load on the tower in kiloNewtons.
Rated Power output @ 11/ms : Given the large sweep area an the more efficient high solidity, high torque design at a much lower TSR of 2.5, Starwind5 conservatively delivers 28% of nameplate at the lower height compared to Bergey which delivers 22% of Nameplate vs UGE 11% of Nameplate (4kW NP used?)
The important overall number though is what is the LCOE Levelized Cost of Energy based on the AEP Annual Energy Produced. We will save that discussion for another Blog shortly, once we have the competitive information collected from trusted third part sources.
Certification: UGE When? still Work in Progress? for 3+ years, hmmm, yes we are working to get this done too, by end of 2016, paid for with licensing fees from our 4 partners.
What next? Starwind5 Production, Deployment, Field Results and Certification
Moving forward, we will continue to learn from the experts and our competitors, and most of all our customers and integrators as we install and monitor the results of the first few Starwind5 systems in the coming months via our licensed partners. We have paid attention to the expert critics, made some adjustments, tuned our design, done some measurements, and to our own encouragement, the others have been found wanting, so now it’s time to deliver and get certified. 🙂