WHI Harvester VAWTs
Overview and Modeling
Vertical axis wind turbines (VAWTs) have been around as long as horizontal axis wind turbines (HAWTs). Both have their origins in wind mills that turned millstones or pumps centuries before electricity was being generated. By the early 1980s, investments and progress on HAWT and VAWT technology was close to equal. From this investment base came the foundational computer modeling and international certification (IEC 61400) requirements needed for both technologies to succeed.
VAWTs have had a difficult time reaching commercialization. Many gained a negative reputation in large part because of the absence of rigorous field testing, data collection and computer model validation of the turbines before sales began.
Sandia National Labs (SNL) in the U.S. produced an important body of work on Darrieus-type prototypes including the Nastran-based Frequency Response (FR) code starting in 1981. Dr. David Malcolm helped with the development while he was at Indal Technologies Inc. and subsequently applied the SNL code to the Indal 6400 (500kW), comparing the code results to field data with some success (see references 1,2). Later (beginning in 1990) Malcolm developed an equivalent PC-based code (the “Eole suite”) which did not require a main-frame computer and expensive license. This code has been largely validated against the SNL code for natural frequencies and forced response (see ref. 3).
The SNL LIFE code is an add-on to either field data results or to FR code predictions. It takes the rms (a measure of the cyclic signal) of the critical load or stress and predicts the likely fatigue cycles, and hence the fatigue life of the component. The LIFE code Malcolm used for the WHI Harvester VAWT was a version started from the SNL code, which then evolved independently.
Field validation with the Harvester 1.0 of these two computer models along with the Midas NFX Finite Element Analysis model were critical to WHI’s engineering and design of its 70 kW VAWTs. WHI’s engineering team used these models and prototype data to finish the Design Evaluation for its Harvester VAWTs and submit these documents to the U.S. Small Wind Certification Council.
Also core to the Harvester’s design were the insights and engineering skill of WHI cofounder and turbine inventor Bob Thomas. He gained a wealth of experience in designing, building and operating 13 prototypes. He and the company learned a great deal about straight-bladed VAWTs, and in the process discovered and patented the “coupled vortex effect” (CVE). However, only WHI’s most recent VAWT prototypes have used strain gauge validated modeling in their design.
Wind Harvest International (WHI) has submitted the Design Evaluation to the U.S Small Wind Certification Council to initiate the certification process of its Harvester VAWT. The next step is to build the v1.1 turbine, install it at the UL and West Texas A&M University’s Advanced Wind Turbine Testing Facility, and put it through the Safety and Function, Power Performance, Acoustic, and six-month Endurance Test. At the same time, UL will install strain gauges on the v1.1. VAWT and produce third-party validated data that can be used to certify WHI’s aeroelastic model, leading to the Harvester becoming the first known VAWT to meet full IEC 61400 certification requirements.
Strain gauge and sensor data collected from the operation of the Harvester v1.0 VAWT prototype in Denmark was key to validating the natural frequency, aerodynamic and fatigue-life computer models that Dr. David Malcolm, WHI’s senior engineer, brought with him from his 30+ year career with wind turbines, along with the Midas finite element model that the Harvester’s lead engineer Antonio Ojeda validated and conducted. With confidence based on the computer modeling and the innovations that emerged from the many years of field testing of WHI’s Windstar VAWTs conducted by Bob Thomas, the Harvester v1.1 is now ready for certification, and the core of an aeroelastic model is ready for validation and certification.
Historical VAWT certification challenge:
WHI knows of only one VAWT that has been certified at a national agency level, the residential 5kW Quiet Revolution. The key challenge VAWT certification traditionally faces is the lack of a fully validated and accepted aeroelastic computer models such as HAWTs have. The model allows the manufacturer to fully account for, and thus properly engineer for, the different loads that the VAWT blades realize as they course through their 360 degree vertically oriented revolution. These loads are substantially different than the loads realized by HAWT blades. Without an aeroelastic model that has been validated with field data, the certification agencies are in uncharted territory as to how to fully certify a VAWT with the same confidence they have with HAWTs.
WHI’s Harvester is a straight-bladed VAWT, which has evolved from earlier versions and modern aerodynamic modeling. It has three aircraft aluminum blades, each 14 m long, and three sets of horizontal arms with a diameter of 12 m which, brings its rotor swept area to 170 square meters with the blade tip fairings have been added.
The Harvester VAWT is cantilevered from the ground through two sets of radial thrust bearings and one set of axial load bearings resulting in a blade tip height of 18.26 m with its shortest 4m tall tower. The rotor drives a direct drive or gearbox driven synchronous generator in conjunction with a power converter that allows variable speed operation. It has a dual, fail-safe braking system using a resistor and a pair of spring-loaded caliper brakes.
The Harvester prototype (v1.0) was field-tested in Denmark at the national Nordic Folkecenter for Renewable Energy. Here is a video of it passing its braking test in high winds. Using strain gauges and other instrumentation, critical data was secured to validate the aeroeleastic modeling used to perfect the Harvester v1.1.
The Harvester v1.1 is designed to operate in IEC Wind Turbine Class II winds and withstand 50-year extreme wind speeds of 59.5m/s (133 mph). With a total weight of 12,000 kg (26,500 lbs), it has a design life of 20+ years with the blades and the core of the turbine engineered to last significantly longer.
Though it will be originally certified with a 70kW capacity generator, it can be used with generators with name plate capacities as low as 50 kW, The choice will depend on the wind resources at the installation site, the Power Purchase Agreement, and governmental and utility policies on project sizes. Customers will be able to choose various height towers (4-8+m tall) to optimize cost effectiveness in different wind shear conditions.
The Harvester is a variable speed turbine, which allows the rotor to operate at maximum aerodynamic efficiency over a range of wind speeds. It also allows limiting the maximum power in higher wind speeds by restricting the rotor speed in combination with aerodynamic stall. The variable speed capability is achieved through the use of a power converter together with a direct drive, synchronous generator.
The relationship between wind speed, rotor speed, and electrical power is shown in Figure 1. Note that while the values in Figure 1 are theoretical, they are based on field data and code validation.
Figure 1: Electric Power and Rotor Speed vs. Wind Speed,
Single Rotor, No Coupled Vortex Effect
Note: The power at high wind speeds is sensitive to the degree of dynamic stall. If field data does not confirm the above power curve, corrections can be achieved by modifying the rotor speed-wind speed relationship.Figure 1.
The projected annual energy production of the Harvester turbine is shown in Table 2. The wind speeds refer to average wind speeds at the site, at the rotor mid-height. The values in the table come from theoretical calculations and are consistent with the power curve in Figure 1.
Table 2. Annual Energy Production
|Average wind speed||Single Harvester VAWT||With Coupled Vortex Effect|
|6.0 m/s (13 mph)||114 MWh||135 MWh|
|7.0 m/s (16 mph)||167 MWh||190 MWh|
|8.0 m/s (18 mph)||219 MWh||252 MWh|
|Cut-in wind speed||4.5 m/s|
|Cut-out wind speed||25 m/s|
|Rated wind speed||14 m/s|
|Reference wind speed (max 10-minute mean)||42.5 m/s|
|Rated power (at 14 m/s)||~70 kW|
|Rated torque||10.3 kN m|
|Rated rotor speed||65 rpm|
|Max rotor speed (over speed)||74 rpm|
Multiple manufacturers will be involved in the production of each Harvester VAWT as WHI will contract this to companies that can best produce the components for different countries and regions. Components will be shipped to a central facility near the installations which will assemble the turbine and ready it for shipping to the project site. In the U.S., Patriot Modular is fabricating and assembling the first pair of Harvesters used in the certification project.
Only a few components are available “off-the-shelf” (e.g. SKF bearings, caliper brakes, fasteners, generator, inverter). The remaining pieces are custom-made. The costliest components include the central mast, composed of three 22-inch diameter pipes with forged flanges attached to each end by full penetration welds, and the nearly 3 meter-long forged and machined drive shaft to which the central mast connects.
The blades are extruded by Nedal Aluminum in Holland. The fiberglass fairings are handmade by Superior Glass in Oregon. The steel blade arms and lattice base towers can be produced by any quality fabrication facility.
By assembling most of the turbine in the off-site, it can be installed and ready for shakedown testing within two days after delivery to the project site. The base tower acts as the VAWT’s nacelle with the driveshaft installed and connected to the generator and all power electronics, controls and brakes installed before delivery. The fully covered base tower is then hoisted onto a “lowboy” semi-trailer for local delivery.
The pre-assembled base tower/nacelle is lifted from the truck by a crane and set onto its foundation anchors and bolted down. Then the central mast is lifted and bolted to the drive shaft. The blade arms are installed by two workers in a bucket lift. The blades are then lifted by the crane and attached to the blade arms by the workers in the bucket lift. The workers then attach the fairings to the blade arms. Last, the turbines' electrical and communication is connected to pre-installed power and fiber-optic lines lines, and the VAWTs are ready for pre-operational testing.
The design and engineering of the WHI Harvester VAWT evolved directly from the combination of modern computer modeling paired with field-tested results from WHI’s utility-scale prototypes installed in California, Finland and Denmark over the past 10+ years. Many key mechanical components featured in the Harvester VAWTs were born from earlier small-scale prototype turbines designed, built and tested by co-founder Bob Thomas in California over the course of ~30 years of building VAWTs. Collectively, these prototypes operated for over 40,000 hours, with the majority of early turbines installed in the intense wind conditions found in California’s San Gorgonio Pass.
Many of the key lessons learned from the designs of Bob Thomas are incorporated in the Harvester. Computer modeling funded by the California Energy Commission and completed by Iopara Inc. showed that WHI could use lower-solidity rotors in its VAWTs and still create the coupled vortex effect. Field-testing of WHI's 636G-3 prototype (Finland, 2013-2014) proved that a design using only three aligned blades would not produce a problematic torque ripple, and, in addition,that lower-solidity VAWTs would realize a higher efficiency (Cp max).
The biggest change incorporated into the Harvester was the H-type design based on the aerodynamic modeling done by WHI engineer Charliaos (Haris) Kotsarinis and the cantilevered mechanical engineering work of WHI’s Ionut Munteanu. Until this H-type iteration, first field tested in Denmark in 2015-2016, WHI’s prototypes all had external support structures to hold the bearing at the top of the rotor mast in place, either with guy-wires or with arms to aerodynamically shaped stators. Changing to a cantilevered design required a significant increase in the size and length of the drive shaft and in the sophistication and strength of the bearings.
The stories of each of the main prototypes can be found on their individual pages. WHI is especially thankful to all the hard and ingenious work put into their designs by Bob Thomas and his son Dean Thomas, without whom there would be no Harvester VAWT nor the discovery and proof of the coupled vortex effect.