Final Project Report

DE-EE0004573
Whitestone Poncelet RISEC Project

Whitestone Power and Communications
FY2011
Final Project Report
Project Title: Whitestone Poncelet RISEC Project

Covering Period: October 1, 2010 to September 30, 2011
Date of Report: September 23, 2011
Recipient: Whitestone Power and Communications

Award Number: DE-EE0004573
Working Partners: Hasz Consulting, LLC; CE2 Engineers; Energetic Drives, LLC;
Applied Power and Control
Cost-Sharing Partners: Hasz Consulting, LLC
Contacts: John R. Hasz, President, 907-895-4770
jrhasz@haszconsulting.com
DOE Project Team: DOE HQ Program Manager – Jacques Beaudry-Losique

DOE Field Contract Officer – Pam Brodie
DOE Field Contract Specialist – Jane Sanders
DOE Field Project Officer – Tim Ramsey
DOE/NAVARRO Project Monitor – Samantha Quinn
Page 1 of 51
DE-EE0004573
FY2011
Table Of Contents___________________________
Executive Summary ......................................................................................................................3

Project Objectives .........................................................................................................................5
Design Paradigm ...........................................................................................................................7
Components Outline ...................................................................................................................11
Float/Craft ..............................................................................................................................12
Decking ..................................................................................................................................13
Anchoring ..............................................................................................................................19
Turbine/Transmission ............................................................................................................25
Power Generation/Conditioning ............................................................................................45
Electrics/Controls/Monitoring ...............................................................................................47
Collaboration...............................................................................................................................50
Conclusion ..................................................................................................................................50
Page 2 of 51
DE-EE0004573
FY2011
Executive Summary
The efforts of this project were primarily devoted toward developing a practical River In-Stream
Energy Conversion (RISEC) device for Alaskan rivers. This resulted in several important
benefits to RISEC research specifically, and for alternative energy research in general.
1. Surveying and site analysis - This project contributed significantly toward determining
suitable sites for RISEC application in Alaska, as well as providing precisely surveyed
maps of the area.
2. Theoretical Modeling - The project also contributed theoretical models for all structural
components. These models were thoroughly analyzed using analytical closed-form
equations as well as finite element analysis. Additionally, kinetic flux and power output
calculations were applied and validated.
3. Prototyping and Experimentation - Several important components, notably turbine
blade and mounting components were prototyped and tested. These tests validated
analytical predictions; resulted in refined, broadly applicable engineered solutions; and
contributed to a cohesive body of knowledge regarding RISEC design methodology.
4. Application Paradigms- This project required the formulation of specific strategies
regarding logistics, debris management, craft assembly and deployment, RISEC/grid
interfacing and craft anchoring. Many of these approaches simplified RISEC
implementation across a broad scope of project scenarios.
Effectiveness and Feasibility
Four crucial factors justify the economic and technical applicability of the device as follows:
1. Efficiency Paradigm - This project analyzed attempted applications of RISEC

technology in Alaska, and concluded that two primary factors determine system
efficiency - turbine efficiency and operational up-time. Many turbines with high
theoretical efficiency were investigated, but in the debris laden Alaskan waters, potential
down-time and costly maintenance and repairs prevented meaningful application. This
project formulated a design which combined efficient power extraction with high
robustness. This ensured continuous and consistent output across a wide range of
environmental conditions.
2. Remote Location Application - Economic effectiveness is largely contingent on the
pay-off period of an installed device. This particular device is designed for Alaskan
villages, which may have kilowatt-hour costs of up to $0.90. A current economic model
for a 100 KW model operating 8 months a year includes a 1.8 million dollar project cost
covering component cost, assembly and installation. An average load of 100 kW at
$0.90/kWh equates to $259,200 annually. A $1,800,000 installation will then yield a
simplified return on investment of 7 years. Each installation is anticipated to function for
30 years, which would mean an average power cost of approximately $0.21/kWh. The
project return on investment would be prohibitively long for locations with ready access
to inexpensive power; however this installation is readily justifiable for application in
Page 3 of 51
DE-EE0004573
FY2011
remote Alaskan villages. It is worth noting that the entire project cost is comparable to
the price of a new diesel power plant.
3. Component Methodology - An important factor in providing efficient and low cost
power was the design choice to integrate Commercial Off-The-Shelf (COTS) technology
into the design wherever possible. This project integrated stock items for mechanical
craft components such as pontoons, transmission, generator, connection and anchoring
hardware. Additionally, the project integrated a novel electrical control system designed
by Energetic Drives, LLC. This system integrated stock electrical components to provide
efficient and clean power output, optimal turbine performance, and operational
versatility. The choice to employ commercially available technology was beneficial for
three reasons. First, the time devoted to designing new components was reduced,
allowing more time to meaningful application research. Additionally, proper application
of state of the art technology improved overall product performance. Finally, installation
and replacement time and cost was reduced.
4. Permitting - Many novel concepts in RISEC technology have been discussed; however,
many designs require permanent structures or involve disturbance of the riverbed and/or
significant alteration of wildlife habitats. While these devices may eventually be
successfully permitted, such design choices imply extensive permitting efforts. In
contrast, this project involved closely working with permitting agencies to specifically
engineer a design with streamlined and realistic permitting goals. While requiring
adherence to strict design constraints, the resulting environmentally friendly design will
ultimately pay off by reducing permitting time at each deployment site.
Page 4 of 51
DE-EE0004573
FY2011
Project Objectives
a) A feasibility study that describes the basic properties and operational characteristics of the
technology, and identifies the technical and economic merits of the concept (TRL 1-2)
i) Completed prior to the start of the project
b) A preliminary design and engineering (TRL 1-2)
c) A systems engineering analysis that may include a needs analysis, requirements flowdown to define
R&D pathways, work breakdown structure, concept definition, management plan, and risk
assessment (TRL 1-2)
i) Completed during the second quarter of the grant period
d) Consider and identify potential deployment sites and the associated potential resource
e) Identification of the intended marine resource application, with potential extractable energy
estimates
f) Engineering and design focused on advancing the device/component for proof of concept modeling,
developing solutions to technology hurdles, determining all components/subsystems, developing high
fidelity estimates of such values as device/component size, weight, layout, interfacing and
performance (TRL 3)
g) Small scale prototyping and testing of components to reduce uncertainty provide input into numeric
models and validate high level assumptions (TRL 3)
i) Planned Work for the Quarter: It was planned to complete this task during the fourth quarter.
ii) Actual Work Completed During the Quarter: This task was partially complete when the
project began. The gearbox transmission, permanent magnet generator, electronic controls
systems, floatation systems, anchoring systems and propulsion systems were individually tested
to the satisfaction of the WPC technical team. However, the design of the prime mover wheel and
the blades which engage with the water required more resources than originally planned. As a
result, the process of prototyping this component and initiating the completion of a scale model of
the entire system was delayed until the fourth quarter.
iii) Explanation of Variance: The design of the blades changed substantially from the conceptual
design model delineated in the conceptual design report (CDR) submitted with the initial
application to DOE. For this reason, prototyping was delayed till the fourth quarter. All
component prototyping has now been completed.
h) Assess Commercial Off The Shelf (COTS) equipment that can be employed within the system
i) Develop specifications for a proof-of-concept model and fabrication plan/costing
j) Test and integration plan
i) Completed during the third quarter of the grant period
k) Numerical model(s) and simulation(s)
i) Planned Work for the Quarter: It was planned to complete this task during the fourth quarter.
ii) Actual Work Completed During the Quarter: This task was completed in full during the
quarter.
iii) Explanation of Variance: Due to a greater research burden than originally anticipated, the
design of the blades for the prime mover was not ready for prototyping as quickly as planned.
That process is now complete with the result of full validation of the theoretical design.
Page 5 of 51
DE-EE0004573
FY2011
l) Assessment of risks and barriers - resource, environmental, ecological, stakeholder, etc. Define a
proposed follow-on RD&D effort that seeks to prove out the concept
i) Completed during the third quarter of the grant period
m) Conduct stage transition design reviews (go/no-go commitment criteria)
n) Consider and discuss Permitting and NEPA requirements where needed to meet future testing and
deployment plans.
Page 6 of 51
DE-EE0004573
FY2011
Design Paradigm
Overall Design Criteria

Alaskan river environments and permitting requirements have placed significant design
constraints on hydrokinetic turbine development to date. Alaskan rivers tend to be swift, shallow
and debris laden; and many potential areas for hydro power development are not readily
accessible. Additionally, many rivers and streams are sensitive and environmentally significant
habitats. This incurs significant challenges for RISEC development; consequently, the following
stringent design criteria were developed:
Environmental Criteria
1. Turbine shall not disturb the river bed, incur risks of pollution, or harm either land or
aquatic habitats.
2. The system shall not involve any permanent structures.
3. The turbine shall not require impoundments or races to constrict or substantially alter
water flow.
Assembly Criteria
1. Turbine design shall be modular; turbines shall be easily specified and outfitted for a
wide range of remote locations and power needs.
2. Turbine assembly and deployment shall be readily accomplished in remote locations
without requiring on-site welding or machining.
3. All components shall be sized for easy shipping to any potential deployment location.
Performance Criteria
1. Turbine must be able to produce power over a wide range of river height and velocity
levels, and withstand high debris load flows.
2. Turbine shall have simple mechanical operation and low maintenance effort and costs.
3. Turbine shall be able to function consistently to provide standalone power, provide power
cooperatively in tandem with one or more power sources, and provide power on an
infinite grid.
Engineered Solutions
Modern research in hydrokinetic technology has typically focused on axial flow turbines such as
Darrieus turbines, and vertical cross flow turbines such as Grashov or Kaplan turbines.
Considerable research effort has been devoted to improving the coefficient of performance
(power output/power available) for these turbines. These turbines typically turn at comparatively
high RPM (60-100 RPM).
Page 7 of 51
DE-EE0004573
FY2011
Most of the turbines above share the common drawback of requiring rapid rotation for efficient
power production. Such lightly built turbine assemblies provide low torque, high speed
operation, and are efficient when running. However, they remain vulnerable to debris collision.
Additionally, full submersion demands a deployment depth no less than turbine height, and
potentially threatens aquatic life.
The solution considered here was an undershot cross flow turbine. The Poncelet style turbine
extracts optimum energy when blade tips travel at 40% water speed, implying a high torque, low
speed turbine. This would require a transmission for practical electric generation. It allowed for
deployment in shallow water and a robust design attenuated problems encountered with debris
collision. This design was eventually chosen for the project.
Consequently the general design paradigm was as follows: The craft would consist of two
pontoons supporting a deck. On this would be mounted the Poncelet style turbine, a
transmission, an electrical generator, and any controls and electrical components. This assembly
would be positioned in the current using an anchoring system entirely fastened to the river bank.
Paradigm Shifts
Within the context of the engineering solution above, several significant paths of design
methodology were considered during development.
Figure 1: Early embodiment of RISEC device
Page 8 of 51
DE-EE0004573
FY2011
Navigation and Deployment
Initial design concepts included a pilot station, controls, and motor on the craft itself. A self-
propelled craft was certainly the most compact method of deployment, and eliminated the need
for another vessel for moving the craft. However, this method required installing control, fuel
storage, and a motor which would be used infrequently, and could not be otherwise utilized.
Additionally, expensive and unmonitored components of this nature in a remote environment
might increase the incidence of vandalism or theft.
For this reason, a paradigm shift was made toward utilizing another boat to deploy the craft. To
this end, a workboat with "pushing knees" was specified. This boat would be capable of pulling
or pushing the craft into position, and would additionally be useful for transporting workers,
tools, and components to and from the craft. The boat would be secured to the craft using a
cabling system. In many remote communities, such boats are likely to be available to be rented
for the project allaying any need to purchase additional hardware.
Figure 2: Early Embodiment of RISEC device
Power Generation
The initial design involved using a compact, inexpensive induction generator for power. This
paradigm involved low costs for the generator, but implied certain design constraints. For
instance, an inductive generator required excitation to produce power, and had specific
synchronous speeds it must exceed before it would produce power. This meant that certain
mechanical braking controls would be installed. Additionally, an induction generator could not
Page 9 of 51
DE-EE0004573
FY2011
be relied upon to provide standalone power. Essentially, the initial design would be more
compact and inexpensive, but less versatile in application.
A number of factors contributed to a significant paradigm shift regarding power generation. It

was desirable to design a craft able to provide standalone power, interface with other power
generation sources, and provide power to an infinite grid. Additionally, a search was conducted
for a more efficient solution for providing clean power; this led to the discovery of and
collaboration with Energetic Drives, LLC. The benefits in terms of power generation and
mechanical simplification caused a significant paradigm shift; the final model had a more
expensive and heavier permanent magnet generator. It was anticipated that, on balance, the
benefits from mechanical simplification, efficiency, and versatility, would outweigh the costs of
a permanent magnet generator.
Figure 3: Final embodiment of RISEC device
Page 10 of 51
DE-EE0004573
FY2011
Components
Float/ Craft
I. Pontoons
1. General Design Requirements (loading, debris, fastening)
2. Previous Designs (material selection process, manufacturing availability etc.)
3. Advantages of current design
II. Decking
1. General Design Requirements (loading, twisting moment)
2. Design parameters (size, material, section geometry)
3. Advantages of current design
Anchoring
I. Cables
1. General Design Requirements
2. Vortex Shedding
3. Mounting Considerations (pulleys, height adjustments)
4. Debris (Shedding, deflection etc.)
II. Rigid Strut
4. General Design Requirements
a. Buckling Load
b. Vertical Load
c. Assembly
d. Water level variation
5. Previous Designs
a. Monopole
b. Sliding Unit Types
c. Fastening Types
6. Advantages of Current Design
Turbine/ Transmission
I. Blade Design
1. Previous Designs.
2. Geometry (dictated by Poncelet)
3. Materials (dictated by geometry <machinability issues>, loading)
4. FEA, analytical, experimental results
II. Turbine Section Design
1. Design Requirements (modularity, simplicity, loading, etc.)
2. Previous Designs (axle/spoke, etc.)
3. Materials
III. Bearings/adjustments
1. Design Requirements
2. Methods (screw jack, ball screw actuators etc.)
IV. Transmission
1. Design Requirements (required gear reduction ratios)
Page 11 of 51
DE-EE0004573
FY2011
2. Previous Designs (chain, belt drives etc.)
3. Benefits of Brevini two-stage epicyclic (coupling options, maintenance etc.)
Power Generation/ Conditioning

I. Generator
1. Design Requirements (rotation speed, power output, flexibility, cost, weight)
2. Previous Designs (induction generator)
3. Advantages of PM motor
II. Conditioning
1. Design Requirement ( universal grid, stand-alone, diesel pairing)
2. Previous Designs
3. Advantages of Energetic Drives System
Electrics/ Controls/ Monitoring

I. SCADA controls
II. Emergency Alert System
Float/Craft
The craft design was subject to specific operational requirements. In order to maximize stability
and load handling, a pontoon mounted craft was specified.
Pontoons
The pontoon design had several requirements. Pontoons are required to be light, resistant to
debris, tough, and equipped with appropriate fastening hardware.
Initially, an aluminum design was considered. A pontoon with required floatation and weight
was specified; however some concern was voiced that debris collision or dragging along rocky
terrain during launch might dent or permanently deform pontoon skin. Additionally, aluminum
pontoons are comparatively heavy.
Fiberglass pontoons were also investigated. However, fiberglass was considered more likely to
crack or splinter under collision or abrade if dragged over gravel or rocks during deployment.
Having rejected the idea of using fiberglass or aluminum pontoons, the concept of high-density-
polyethylene (HDPE) pontoons was investigated. HDPE has low flexural stiffness; however a
stiffening channel section fastened on top attenuated this problem. It was recommended to fill
the pontoon with closed cell foam to ensure continued floatation in case of hull failure.
Additionally a steel plated pulling head option for cable attachment was offered with a load
capacity of over 200,000 lbs. This exceeded operational requirements for anchoring. An
Page 12 of 51
DE-EE0004573
FY2011
aluminum cone would be installed over the pulling head to reduce the energy loss due to the
occurrence of turbulent flow around the pontoons.
The advantages of this design were numerous. The resulting pontoons were comparatively light,
structurally sound, tough, relatively inexpensive, and offered robust performance during
deployment and operational phases.
Decking
Due to a combination of design choices, the craft was specified with a turbine mounted with a
generator on one side and a free bearing on the other. During operation, this implied a twisting
moment between the torque (generator mount) side and the non-torque (plain bearing) side of the
craft.
The force distribution through the pontoons, decking, and frame components was complex.
Understanding the forces and designing components to withstand them, was a crucial aspect of
project development.
The torque is transmitted to the frame through the generator mount and exerts a rotational
moment on the pontoon which "buries" the upstream side of the pontoon and lifts the
downstream side. The torque is transmitted through the decking (which is rigidly attached to the
mounting channels on each pontoon) to the plain bearing mount pontoon. Thus the pontoons
share the torque loading of the blades by rotating to equilibrium. Any difference in co-planarity
of the pontoons would be due to distortion in the decking. Additionally, anchor cable placement
implied a compressive axial load in the decking.
First it was desirable to determine what angle of heel the craft would assume due to the torque,
and then assess the internal stresses in the craft frame the moment would create. Assuming a
static equilibrium, it was assumed that the torque moment must be resisted by an equal and
opposite "righting moment".
This may be related to the angle of heel by the following equation:
R.M .    Volumedisplaced  Height metacentric  
where α is the angle of heel, in radians, γ is the density of water, and metacentric height the
distance between the metacenter and center of gravity of the craft.
To determine values for substitution, first the center of gravity was calculated. For a given
number of objects with known heights and weights, this may be expressed:
Weight1  Height1  Weight 2  Height 2  ...  Weight n  Height n

Weight1  Weight 2  ....  Weight n
A simplified center of gravity for craft + wheel was estimated as follows:

A 15,000 pound craft with CG at 4 feet, 5000 pound wheel with CG at 7 feet.
Page 13 of 51
DE-EE0004573
FY2011
15,000  4  5000  7
 4.75
20,000
Thus a simplified estimate of CG is 4.75 feet.
The center of buoyancy could be readily calculated by determining the CG of displaced water.
This was done by simplifying the model analytically by assuming the minor center of buoyancy
change under loading makes a negligible difference in calculations- this was validated later.
A weight of 20,000 lbs in water with a density of 62.5 pounds per cubic foot required 320 cubic
feet of displacement. To simplify calculations, the pontoons were considered to have a square
rather than round cross section - the difference being assumed negligible (this too was validated
later, as will be seen). For a 34 foot simplified pontoon of 3.5 foot width and height, the
immersion height is 2.68 feet.
Metacentric Height is calculated as follows:
I
Height metacentric   Centergravity  Centerbuoyancy
Volumedisplaced
3
I Widthcraft  Lengthcraft 19  34 3
   194.47 feet
Volume displaced 12  Volume displaced 12  320
It is clear that the primary factor in metacentric height is inertia controlled, so the simplifications
of square pontoons and small heel angle are validated since they have negligible effect- the
metacentric height was then
194.47 - 4.75 + 2.68 = 192.4 feet
Thus the overall craft angle of heel is calculated by substituting into equation above:
R.M . 105,000 ft  lb
   0.027radians  1.56 deg
  Volumedisplaced  Height metacentric lb
62.4 3  320 ft  192.4 ft
3
ft
Since the decking provides the sole structural interface between these elements, the decking must
be sufficiently stiff to withstand this torque. To this end, a decking solution was sought which
would fasten between the pontoons. Such a decking design would need to be lightweight and
resistant to bending and twisting- that is, a high polar and area moment of inertia.
Page 14 of 51
DE-EE0004573
FY2011
Figure 4: Hollow decking section showing cross section with support rib
Hollow, closed, geometric sections were chosen for decking cross-section, since they combine
high area and polar inertial moment with low weight. The decking had an additional design
constraint of being flat- this led to the choice of a hollow rectangular cross section. The area
moment of inertia would be analytically expressed:
Widthouter  Heightouter
3
Widthinner  Heightinner
3

12 12
and bending stresses would be expressed:
Moment  C

I
where C is the distance from neutral axis to outer edge of beam, and I is area moment of inertia
For an axial torque, the maximum shear loading for a thin walled beam is determined by first
calculating shear flow in the hollow section:
Torque
q
2  Areamidline
where q is the shear flow, and midline area is the area defined by the midline of the beam cross
section.
Page 15 of 51
DE-EE0004573
FY2011
Shear stresses are found by dividing the shear flow by wall thickness. Then shear stresses and
bending stresses are combined to determine principal stresses.
2
x y   y 
 a ,b    x    x , y
2  2 
Alternatively, the von Mises effective stress may be calculated in terms of applied stresses:
 '   x2   y2   x y  3 xy2
The problem with this method was that the specific geometry of the craft implied that the highest
bending stresses occurred at the ends, with lower stresses in the middle - since for a given angle
difference between the pontoons, the forces were not equally distributed. It was very difficult to
develop an accurate closed form equation to describe the stresses due to a combination of
bending and twisting.
The design methodology was as follows: the decking would be of uniform height, and would
need to be able to transmit torque between the pontoons without incurring unacceptable stress
levels. The pontoons themselves were considerably less stiff than the mounting channels on top
of them. Thus the mounting channels were designed to maintain shape and integrity under axial
twisting and transverse bending loads - the small displacements were not anticipated to produce
high stresses in the pontoons (see FEA results in figures 5, 6, and 8).
Material selection was an important design decision. Steel was considered for its ease of
welding and construction, and high strength. However, certain aluminum alloys offered superior
strength to weight ratios and better corrosion resistance. Below is a table of relevant mechanical
properties for several candidate materials.
Material Properties Table

Material Type Density Elastic Yield Strength Fatigue
(lb/in) Modulus (psi) (psi) Strength1 (psi)
440 C annealed 0.28 30,000,000 65,000 33,000
Stainless Steel
304 annealed 0.28 30,000,000 35,000 17,000
Stainless Steel
5086-H32 0.10 10,300,000 33,000 23,000
Aluminum
5086-T0 0.l0 10,200,000 17,000 N/A
Aluminum
7075-T6 0.10 10,400,000 83,000 23,000
Aluminum
6061-T6 0.10 10,000,000 40,000 14,000
Aluminum
6061-T0 0.10 10,000,000 8,000 9000
Aluminum
1
At 500,000,000 cycles- measured in fully reversed bending using R.R. Moore apparatus and sample type
Page 16 of 51
DE-EE0004573
FY2011
Several noteworthy details with regard to design are shown in the table above. First, heat
treatment and tempering make considerable differences in mechanical properties. Additionally,
high yield strength is not necessarily an indication of fatigue performance. The performance of
candidate materials under repeated load cycling was of definite significance in material choice.
Eventually 5086 (aluminum-magnesium alloy) was chosen for decking construction. The
resulting deck pieces were constructed of hollow rectangular sections with widths varying from
18-24 inches and a height of 8 inches.
Since the individual bending loads in the decking sections were difficult to calculate utilizing
closed form analysis, an FEA model was developed to ensure that the decking and channel
components were sufficient. In this model, a moment was developed at the generator mount
pontoon, and a cable anchoring force at the pulling head of the other pontoon. Both pontoons
were constrained at the ends using theoretical radially flexible spring bearings to simulate water
buoyancy and floatation, and a roller constraint was applied at a decking section to simulate the
rigid strut constraint (more detail about this design element will be presented in a later section).
A limitation of the model is that the spring bearings are an imperfect model of water support in
several ways. First, they constrain in every radial direction rather than merely providing buoyant
forces. This causes artificial resistance to anchor cable force where water floatation opposes
vertical but not horizontal motion. Furthermore, since all bearing was at the ends of the
pontoons, the FEA bearing stresses would be higher than actual stresses. Additionally, the
moment in the model was exerted, not at the generator mount, but at the pontoon (see Figure 5).
Figure 5: FEA stress and deflection plot showing loads and constraints
Nevertheless, the FEA testing resulted in displacements and stress which validated expectations.
Note the deviation from planarity caused by warping in deck elements in figures above and
Page 17 of 51
DE-EE0004573
FY2011
below. The deflections are normalized to make small deflections visible; hence, the aspect ratios
are skewed.
Figure 6: FEA results of decking and frame deflection under operational loading.
To obtain a more precise picture of stress distributions, Solidworks "Iso Clipping" was utilized to
select minimum stress value to display. Note that the only locations above 2000 psi are in the
inner deck plates and mounting channels, with a maximum stress of 5,341.4 psi. For 5086
aluminum, this indicates an acceptable factor of safety of 4.31.
Figure7: ISO clipping- stress <2000 psi
Page 18 of 51
DE-EE0004573
FY2011
To isolate and examine the stresses in the pontoons, a minimum display value of 102 psi was
determined. As anticipated, most stresses in the pontoons were lower than this value. For yield
strength of 3000 psi, this indicates a factor safety of 29.
Figure 8: ISO clipping- stress > 102 psi
Anchoring
Cables
An anchoring system was required to prevent craft motion during deployment. Several
significant environmental factors incurred design requirements on the anchoring system.
1. Permitting constraints required any system to have a small footprint. Any design
including disturbance of, or anchoring to, the river bottom, or involving any permanent
structure would require prohibitive permitting efforts (not to mention that river bottom
profile changes could involve undesirable anchor point motion). Consequently all system
components must be portable, environmentally friendly, and non-invasive of the river
bed. This implied a system anchored to the shore.
2. It was desirable to integrate debris diversion with anchoring systems since such systems
would necessarily bridge between the craft and shore. Although all craft components
were designed with debris collision survivability in mind, the intent of anchoring design
was to eliminate this hazard as much as was practically feasible.
3. An anchoring system was desired which would not substantially hamper flow to the
turbine, or harmfully accrue debris on any individual component.
Page 19 of 51
DE-EE0004573
FY2011
4. Water elevation changes substantially (up to ten feet over the course of a season). This
required any anchoring system to be either adjustable or otherwise configured to provide
support at a variety of water levels.
5. Any anchoring system would require some configuration to prevent the craft from
moving toward the shore. Although apparently self evident, this requirement necessitated
the inclusion of some rigid elements to hold the craft a fixed distance from the river bank.
To satisfy these general requirements, several designs were considered. The preliminary design
was a monopole which was installed perpendicular to the bank. Cables would be installed to
each pontoon. Height adjustment required that the pole have some articulation at its connection
points on the craft and on the shore.
The other potential design plan was to mount a cable from the shore to the opposite front corner
of the craft. This cable would provide an anchoring point and would be run just under the water
surface to provide debris diversion. It was anticipated that large trees, especially those with root
wads, would strike the taught cable and be diverted from craft. This method would imply a
sideways force tending to push the craft toward the bank. Preventing this would require a rigid
strut to maintain position.
Several variants of this design were considered. A preliminary proposal suggested providing
vertical adjustment by mounting all components on dollies such as those used for overhead shop
hoists. These dollies would then be actuated by a servo or crank controlled ball screw system.
The dollies themselves would track on vertically oriented I-beams which would be fastened to
the river bank. This would afford controllable height adjustment varying with river levels.
Another potential design to reduce cable size involved the use of pulleys running through
sheaves attached to the anchor points and craft. A single capstan on the craft would then reel the
cable in or out.
These designs were eventually abandoned in favor of a simpler design. A rigid strut member
would consist of modular suspension bridge segments. These could be individually installed as
suspension bridge segments above the water surface at low water. The connection between the
craft and this suspension strut would be a king-pin/fifth wheel connection such as is employed
for RV or trailer towing. The suspension strut would be fastened to the shore by a custom pintle-
style mount. This would allow it to bear axial loading, and also to tilt to accommodate varying
river levels from a single shore position. The debris diversion cable would be strung from the
shore to the opposite side of the craft. On the shore side, a series of rock anchor tie off points
would be provided. As water levels vary, the cable could be installed at different points to
ensure proper cable depth and functionality.
The actual installation height of the rigid strut relative to high/low water levels was a significant
design consideration. An initial suggestion of splitting the difference between minimum and
maximum water level height was made. However, at the desired installation site, water level
varies approximately ten feet seasonally. At maximum water height (and maximum force in the
strut) a five foot vertical difference over a thirty foot span makes a 9.5 degree angle between
Page 20 of 51
DE-EE0004573
FY2011
strut and water. The vertical force is equal to the axial force in the strut multiplied by the sin of
the angle- for a 10,000-pound strut force, this meant a 1,643-pound vertical force. This force
stresses fifth wheel mounting structure. Considering that the highest forces are at high water, the
bridge anchoring position was changed to 3/4 high water mark to make forces more equalized
over deployment time.
Having determined the general method of securing and anchoring the device, individual
component design was considered. Both the cables and suspension strut components were
designed to be anchored into the shear rock face of the bluffs at the river bank. To this end,
threaded rock anchors were specified for fastening components to the rock face. Manufactured
by Williams Form Engineering, these rock anchors are one inch in diameter and five feet long.
They are grouted into a pre-drilled holes and have a pullout strength of 60,000 lbs in the quartz-
biotite-gneiss rock which comprises the bulk of the bluffs.
The cables were specified with the following design criteria: The cables must be strong enough
to bear the operational loading of the current flow plus any forces set up by debris impacts or
accumulation. A flow of approximately 15 feet per second was calculated to exert a force of
approximately 14,000 lb in the direction of the current. An additional debris impact was
calculated to exert approximately 3000 lb. Thus the total load in direction of the current is
approximately 17,000 lb. If the angle θ between the cable and current direction is approximately
30 degrees, then the actual force in the cable is expressed
F = 1/cos(θ)
= 19,600 lbs
Using simple force equilibrium principles, the resulting force in the rigid strut is expressed:
F = (1/cos(θ))sin(θ)
= 9800 lbs
Several dynamic considerations were made in cable design. Some concern was discussed that at
certain river speeds, vortex shedding frequency from the diversion cable might approach the
cable's natural frequency, causing cable flutter. An equation was derived for the natural
frequency and dynamic behavior of a flexible cylinder under tension with pinned ends.
The natural frequency of a tensioned cable in water may be very closely approximated as
follows2:
1 Tension
fn 
2  Length Massunit _ length water  Massunit _ lengthcable
1 20,000
fn   1.71 Hz
2  1440 0.000731  0.0000937
2
Dauchin, Benoit. Flow Induced Vibrations on a Cable Caused by Waves Plus Current. Diss. Ecole Centrale de Lyon
France, 1996.
Page 21 of 51
DE-EE0004573
FY2011
The vortex shedding frequency is governed by the Strouhal number, a dimensionless parameter
which is itself dependent on the Reynolds number:
Velocity  Diametercable
Re 
Vis cos ity fluid
The Strouhal/Reynolds relation is graphically presented in the following figure:
Figure 9: Chart relating dimensionless parameters for a cylinder in cross-flow
For a velocity of 15 feet per second, the resulting Reynolds number is approximately 65,400.
The Strouhal number is approximately 0.21; this results in a shedding frequency of 42 hertz.
The vortex shedding frequency is described by the following equation:
Strouhal  Velocity
f shed 
Diametercable
Since a 1.71 Hz shedding frequency only occurs at speeds of approximately 1.286 inches per
second, there was no anticipated risk of flutter at operational flow rates.
Rigid Strut
A rigid strut component was required to maintain craft position in current flow. Such a
component was subject to several design constraints. First, modularity was desirable for two
reasons.
Page 22 of 51
DE-EE0004573
FY2011
First, assembly and disassembly was facilitated by employing small, individual subcomponents
which could be installed and uninstalled sequentially rather than handling the entire cumbersome
strut.
Secondly, since variable positioning in the current was potentially critically important, it was
desirable to be able to add or subtract segments to optimize craft placement. The importance of
modularity led to a design specification that the strut be comprised of ten foot sections which
could be fastened together to create a strut of arbitrary length (as long as resulting strut is safe
from buckling).
Obviously buckling failure was a significant design consideration, especially since the strut
would be a long, slender design comprised of several sections. Several methods of analysis were
considered to ensure that material and installation costs were minimized without compromising
buckling resistance. An initial design specification was for six inch diameter 6061 T-6
Aluminum alloy structural tubes with quarter inch wall thickness to be used. Each section would
be comprised of two such tubes placed 30 inches apart on center and cross braced with 1" x
0.125" square tubing.
Since the primary axial loading would be through the kingpin on the bottom of the strut,
eccentric loading was anticipated. An analytical application of the secant method for calculating
critical loading was used first. This formula is expressed as follows3:
Load Strengthcompressive _ yield


Areacross sec tion ec  Lengtheffective Load 
1  sec  
k  k 4  E  Areacross sec tion 
where
2 xI
k
Areacross sec tion
ec is the eccentricity (for this calculation it was assumed to be five inches)
E is the elastic modulus.
Since the closest approximation to realistic end constraints was a pinned-pinned condition, the
effective length was the same as real length, 30 feet.
This was iterated with varying loads until convergence to solution - the critical load was 12,550
lbs for a single 30 foot length. Since each strut section would include two such components,
linear superposition was used to determine total assembled strut buckling load- 25,100 lbs. Since
the anticipated maximum loading (operational forces + debris striking) is approximately 10,000
lbs at the bridge, this provides a factor safety of approximately 2.5. The limitations of this
calculation are that no transverse loading scenario is considered in the secant formula. Some
concerns were discussed that gravitational loading over the 30 foot span, as well as any other
3
Norton, Robert L. Machine Design, and Integrated Approach. Prentice Hall, Saddle River, New Jersey, 2006
Page 23 of 51
DE-EE0004573
FY2011
transverse loading, might significantly decrease buckling resistance. Additionally a higher
moment of inertia was desired to increase buckling factor of safety, thus an 8 inch pipe was
specified for analysis.
To this end, the Solidworks finite element analysis software was employed to create a buckling
study. At this time, the design suggestion had been made to fasten each strut section together
with pipes which would fit tightly on the inside of the eight inch pipes. These would have a
length of 48 inches and would be plug welded on one side to the outer pipe. The other side
would have a hole for a pin, which corresponds to a hole in the outer pipe. This way, each
section may be pinned onto the last, with transverse bending support provided by the inner pipe,
and axial bearing provided by the outer pipe. The FEA model included these inner pipes.
Additionally, the FEA model included the gravitational load, and a transverse load of 500 lb, as
well as an axial load of 5000 lb (with an eccentricity of one foot).
FEA limitations were as follows: the Solidworks package was not able to calculate differences in
buckling/bending resistance at the joints due to pipe clearances; therefore all touching surfaces
were assumed in bonded contact. Additionally the actual design included a kingpin which gave a
further pinned degree of freedom perpendicular to the bearing constraint. This was not included
in the model.
The Solidworks package utilized an eigenvalue calculation to predict buckling shape and
occurrence; this resulted in a loading safety factor of 8.4. This was determined to be acceptable;
thus a final design decision was made to create individual bridge sections from ten foot sections
of 8” x 0.25" structural tubing, and pin each section together with 48 inch connecting tubes.
Figure 10: FEA model of rigid struts
Page 24 of 51
DE-EE0004573
FY2011
Figure11: Results of FEA model
Turbine/ Transmission
Blade Design
A number of requirements were imposed upon turbine blade design. In terms of size, the ratio of
radial length to turbine diameter was fixed for optimum efficiency. Additionally, it was
desirable to design a blade that would prove survivable and robust under operational conditions
and debris strikes. It was also desirable to make any mechanisms highly robust to withstand
submersion in silty water, and reduce moving parts as much as possible to decrease
manufacturing and assembly costs, as well as maintenance.
Geometrically, a curved profile to trap water was desired to increase efficiency. Initial blade
designs were also tapered to save material costs and weight while maintaining constant stress in
the blade. This concept was eventually abandoned due to manufacturing constraints. A primary
design concern was the collision of a log or piece of debris with the blade; a number of potential
designs were considered.
Blade Design Calculations
The power developed by an undershot waterwheel in unconfined flow is expressed4:

vu (v  u )
Fu  BS
g
where F is the force in pounds developed by the water on the blades of the wheel, B is a constant
determined experimentally to be 0.8, S is the total surface area of the blades in the water in
square feet, v is the velocity of the water in feet per second, u is the tip velocity of the blades in
4
Bresse, Jacques Antoine Charles, Water Wheels or Hydraulic Motors, University Press of the Pacific, 2003
(reprinted from 1876 edition)
Page 25 of 51
DE-EE0004573
FY2011
feet per second, π is the weight of water equal to 64 pounds per cubic foot, and g is the
acceleration of gravity equal to 32.2 ft / sec 2 . For the case of wheels in unconfined flow, the
maximum efficiency is obtained when
u
 0.4 .
v
The force developed against the wheel could be determined by dividing the above equation by
the tip speed of the blades u. In the case of this design, 9 blades were considered to be in the
water at one time, with each blade having a total area of 8 ft 2 perpendicular to the direction of
flow. The velocity of the wheel could be considered constant due to the high gear ratio between
the wheel and generator. The electronic controls would use the generator to hold the optimal
u
speed ratio between the wheel and the water of  0.4 at all times regardless of water velocity.
v
Experimental results indicated that the depth of the blades should be less than or equal to ¼ of
the wheel radius. In addition, experiment dictated that for a wheel 16 ft in diameter, the number
of blades should be 12.
The curvature of the blades was determined by the water flow regime and was optimized to
minimize shock as the blades entered and exited the fluid. In addition, the curvature allowed the
blades to absorb more energy than they would otherwise do by lifting the water as the wheel
turns. The theoretical efficiency of such a wheel in a confined flow is 100%, however the
maximum attainable efficiency given friction and fluid escape was somewhat less than 60%. The
curvature of the blades was determined by the approach of the blade into the water and the angle
of the root of the blade to the circumference of the wheel. Experimental results showed that the
approach angle of the blade to the water should be 30 degrees and that the root of the blade
should be perpendicular to the circumference of the wheel.
The total force developed on the wheel by the water at 8 ft/sec is 4,500 lb which computed to
approximately 31,500 ft-lb of torque. At 15 ft/sec the total force developed on the wheel by the
water is 15,500 lb which corresponded to a torque of 108,500 ft-lb. At these water speeds, the
wheel produced 17 kW and 107 kW respectively. This power output took into account the
inefficiencies of the drive train, generator, and inverter equipment.
Blade Design Process
The first embodiment of the above criteria was a curved plastic or aluminum plate with
aluminum support ribs. This assembly would be pinned at the top with a coil spring. A heavy
log strike would cause the assembly to rotate about the pin, with the spring absorbing energy
from the moving log and allowing it to pass under the turbine. This concept was abandoned due
to concerns that the aluminum ribs would permanently distort and that the coil spring would be
constantly in angular displacement under operational load. Additionally, the coil springs would
add prohibitive cost, weight, and installation difficulty.
Page 26 of 51
DE-EE0004573
FY2011
Figure 12: Early blade design with aluminum ribs
A second embodiment was to utilize a detente notch in the side of a metal disk and roller under
spring pressure to hold the blade in place during normal operation. In the event of a log strike,
the collision would cause the ball to pop out of the detente and the blade would rotate out of the
way. When the blade was raised out of the water, gravity would cause it to rotate back to detente
position. Potential designs were generated using Belleville washers, leaf springs, cantilever
springs, and helical springs to provide the force to secure the roller in the detente notch.
Figure 13: Detail of paddle with detente notch mount
Page 27 of 51
DE-EE0004573
FY2011
Figure 14: Early all-plastic blade design
For a detente disk with radius r, and notch angle 2 x θ, the roller would only move in the notch if
force in the spring F was equal to upward force on roller. This may be expressed:
Torque / r
 Fspring
1  cos 2 ( )
A third embodiment considered was the result of concerns that logs and other debris might not
only damage the blades in the event of a direct collision, but also might become pinned under the
turbine and exert a radial force against the outer tip of the blades. A turbine assembly that was
robust to tangential impact forces as well as radial forces required to push a log under water was
desired. To this end, a design employing a rod and coil spring which could move radially inside
a sleeve guide on the wheel spoke was created. This would soften the impact of pushing a log
under water. Ultimately this design was abandoned for three reasons. First, it involved
numerous small moving parts, prompting concerns about silting and corrosion, as well as high
maintenance and installation costs. Secondly, when the spring was compressed, the blade was
still subjected to full buoyant force of the log. Finally, an unrelated design constraint caused the
spoked wheel design to be eliminated, making the guide sleeve and spoke method unwieldy in
the design context.
Page 28 of 51
DE-EE0004573
FY2011
Essentially, all three mechanical methods of reducing log impacts were abandoned because they
required numerous moving components, implying high assembly and maintenance costs. This
prompted a change in design paradigm away from mechanically actuated blade protection
assemblies in favor of a simply fastened but robust blade design. This required a more careful
selection of available materials.
Initial design ideas had relied heavily on mechanical breakaway safety mechanisms to reduce
impact loading on the blades. Abandoning these mechanisms meant that the blade would need to
be able to undergo full impact loading in both tangential and radial directions. Early designs
included fully aluminum blades, and plastic blades with aluminum ribs. While these designs
were typically able to handle operational loads, it was anticipated that debris collisions would
permanently deform metal components.
Consequently, considerable research was conducted to determine a suitable material which

would be both reasonably light and inexpensive, strong enough to hold shape under operational
loading, and flexible enough to bend without permanent set in the event of a log strike.
Aluminum was considered for its machinability and corrosion resistance. However a design
strong enough to withstand a log strike would require a prohibitively high material weight and
cost.
A number of engineering plastics were available, ranging from acetals5 (yield strength: 10,200
psi, elastic modulus: 435,000 psi), polyethylene terephthalate, ultra high molecular weight
polyethylene6 (yield strength: 3100 psi, elastic modulus: 100,000 psi), polyimides, ABS plastics,
and high density polyethylene. Since the simplified blade design would need to be flexible
enough to recover after the potentially substantial distortion of a log strike, percent elongation
was evaluated along with yield strength and elastic modulus for candidate materials. Besides
mechanical properties, a suitable material would also need to be inexpensive in terms of material
and machining costs. These requirements eliminated polyimides and acetals and PET because of
the limited machining options and high manufacturing and material costs. Additionally these
materials had limited flexibility (45% elongation of acetal copolymer and 7% elongation of
polyimide). ABS plastics were more easily machined, but had unsuitable mechanical properties.
Ultra High Molecular Weight Polyethylenes had similar properties to HDPE, with a slightly
lower modulus of elasticity. This meant that it would provide a more flexible blade; this was
initially attractive, but UHMW cannot be welded, which severely limited its application. High
density polyethylene was eventually selected as the material for final blade design. The
requirement of a 20 degree bend in the blade profile required forming that was unavailable for
any other engineering plastic. Especially notable is its 500% elongation at rupture, and its high
degree of shape recovery after distortion. HDPE7 has an elastic modulus of approximately
175,000 psi and an ultimate tensile strength of approximately 3500 psi.
5
Dupont Delrin Acetal Resin- Product Property Guide- 2010
6
Ultra High Molecular Weight Polyethylene (UHMWPE) Harvey L. Stein, PE. Reprinted from Engineered
Materials Handbook Volume 2: Engineering Plastics, 1999
7
HDPE Data Sheet, Chevron Phillips Chemical Company, 2009
Page 29 of 51
DE-EE0004573
FY2011
Below is a table with relevant mechanical properties of some of the candidate engineering
plastics.
Material Elastic Modulus Yield Strength % Elongation at Machinability

(psi) (psi) Rupture
Acetals 435,000 10,300 35 Stamped, extruded,
small parts only
UHMW 100,000 3100 350 Small sheet
extrusion, no
bending
HDPE 175,000 3000 500 Sheets may be
drape formed,
bent, and welded
Analytical Predictions
The following analytical method was employed to determine the stresses due to standard
operation and potential debris strike:
A turbine blade-debris collision was modeled analytically as a case of horizontal striking impact
using the kinetic energy method. The system was simplified by considering the log or debris as a
moving mass with kinetic energy and the blade as an elastic member. In a collision, if we
assume dissipation to be negligible*, all kinetic energy from moving mass (log) is converted to
elastic energy stored in the struck member (turbine blade). This may be expressed:
2
Force 2 masslogVelocitylog

2k paddle 2
where k is the stiffness of the paddle, and η is the internal dissipation of kinetic energy in the
blade (where a value of zero would imply total dissipation, and one no dissipation)*
which may be further reduced:
Force  Velocitylog   masslog k paddle
The stiffness of the paddle (force/displacement) was determined by modeling the paddle as a
cantilever beam with a cross-sectional profile similar to that of paddle. The displacement under
load for a cantilever beam is expressed:
FL3

3EI
where δ is the displacement, F the force, L the beam length, E modulus of elasticity of beam
material, and I the area moment of inertia. Rearranging for force/displacement, stiffness may be
expressed:
Page 30 of 51
DE-EE0004573
FY2011
3EI
k
L3
The area moment of inertia of the paddle was estimated by considering a simplified cross section
consisting of a composite of three rectangles as shown in Figure 11.
According to the parallel axis theorem, linear superposition of these individual beams may be
used to express the moment of inertia as follows:
b1h13 b2 h23 b3 h33

I  
12 12 12
1
Roark and Young provide a factor of correction value8  
massblade
1
masslog
A conservative (and fairly accurate for large logs) simplification is to assume η = 1. Employing
the equations above, the following assumptions were made:
lbs  sec 2
- A log with weight of 1000 lbs, or a mass of 2.6 and an absolute speed of 10 feet per
in
second (120 inches per second).
- A high density polyethylene ( elastic modulus = 100,000 psi, yield strength in tension = 3000
psi) paddle with cross sectional dimensions as shown in figure 10 velocity 40% that of flowing
water, or 4 feet per second.
The area moment of inertia was calculated to be:

1x63 1x6 3 48 x13
I    18in 4  18in 4  4in 4  40in 4
12 12 12
The stiffness was then expressed:

lb
3x100,000 2 x 40in 4
3EI in lb
k 3  3
 781
L 13824 in in
These values may be substituted to determine the maximum force:
in
Force  Velocitylog masslog k paddle  72 x 1x 2.6 x781 = 3240 lb
sec
To determine maximum stress in paddle, the following equation is employed:

Mc

I
8
R.J. Roark and W.C. Young, Formulas for Stress and Strain. 6th ed. McGraw-Hill: New York, 1989
Page 31 of 51
DE-EE0004573
FY2011
where σ is the bending stress, M is the moment, c is the distance from neutral axis to outer edge
of beam, and I is the moment of inertia. This implied that the maximum stress would be found in
those parts of the beam which were furthest from the neutral axis. In this case, 3 inches was the
furthest distance.
Substituting these values:

3240lbx 24inx3in
 max   5832 psi
40in 4
This particular analytical method of ascertaining stress implied several limitations. First, as seen
in Figure 15, the model geometry differed from actual profile in two important ways. First the
analytical model was a straight beam, while the actual profile was a curved blade. Secondly, the
actual beam tapered, whereas the analytical model was of constant cross-section.
The first error caused the model to be less stiff than a more accurate representation would
suggest. The second error caused the model to be stiffer than a more accurate representation
would suggest.
Since the analytical model exhibited considerably smaller deflections with higher stresses than
either experimental or FEA results, it was concluded that the analytical model required
refinement to represent a less stiff blade with lower stresses.
Figure 15: Comparison closed form model to actual model
FEA Predictions
A solid model and finite element mesh equivalent of the blade prototype was generated using the
solid model/finite element software package Autodesk Inventor. The native finite element mesh
generator employed elements with an average size of 0.1 inches, with automatic detection and
Page 32 of 51
DE-EE0004573
FY2011
recalculation of element groups with poor aspect ratios. This allowed very close geometric
tolerances.
A static test was conducted with the following parameters:

1. Material- blade constructed of thermoplastic resin
2. Constraints- fixed constraints on mounting surfaces (see Figure 15)
3. Loads- Pressure load (operational water load) on turbine faces, 3000 lb
force at middle of blade tip (log strike impact force).
The maximum stresses were found in the support ribs at the sides of the blade, as predicted by
the analytical formula. However, the maximum stress was considerably lower and the
displacement higher than predicted by the analytical method. This was concluded to be the result
of limitations in the analytical method; specifically that it did not account for lowered stiffness
due to tapered profile.
Figure 16: FEA results of paddle loading
Page 33 of 51
DE-EE0004573
FY2011
Experimental Testing
Two manufacturers built prototype turbine blades based upon the design requirements and
drawings. A steel fixture was constructed with two primary design considerations. First its
fastening system resembled the actual fastening brackets as closely as possible; and secondly its
spring system made static force calculations feasible. The experimental testing regimen was
two-fold. First it was desired to validate the FEA predictions of load response and survivability
by observing behavior under known loads. Additionally, it was desired to validate analytical
predictions of forces generated by placing the turbine blade in moving water.
Figure 17: Testing jig with springs
The first test was conducted by securing the blade in the steel jig-shown above, which was itself
securely constrained. Then a spring with known stiffness (3000 lb/in) with a steel end was
pressed against the paddle at the tip, and the spring deflection measured. This test was applied to
both paddles to simulate a log strike.
Page 34 of 51
DE-EE0004573
FY2011
Figure 18: Log strike test- note cracking at weld seam.
Figure 19: Blade deformation recovery after log strike test
Page 35 of 51
DE-EE0004573
FY2011
One paddle failed under the load, and one survived. Both blades exhibited excellent recovery
from the distortion- see figure above.
The second test was performed on both blades; in this case the jig was fastened to a metal beam
with U-bolts (see figure below) and the blade secured in the jig; the blade was placed into
flowing water. The forces generated by the flowing water were calculated as follows:
Force  Ycp
 Rspring
l
where Ycp is the center of pressure l is the distance from the pin to the spring, and R is the
reaction force at the spring.
Figure 20: Water flow test setup
Page 36 of 51
DE-EE0004573
FY2011
Figure 21: Blade under water flow test
Figure22: Deformed spring under water flow test
Page 37 of 51
DE-EE0004573
FY2011
Conclusions
After experimental testing was concluded, the final design was a turbine blade constructed of one
inch thick HDPE plate. This was bent and groove-welded to form a curve as specified, and one
inch thick ribs were groove-welded onto the sides. In experimental testing, these blades were
both able to sustain tip displacements of at least five inches, and tip loading of at least three
thousand pounds, although a crack did appear at a welded seam of one of the blades when
loaded.
Turbine Section Design
The turbine section itself underwent considerable design changes. Initially, a spoked wheel
design with a central axle was considered. The initial design incorporated spokes which
transferred torque from the blades to the central axle, which also supported the transverse load of
the total turbine weight.
Figure 23: Early turbine and blade assembly
An initial design specification of an 8 inch diameter axle with 0.75 inch wall thickness and raised
bosses for spoke attachment was suggested. An FEA calculation was made upon this model.
The FEA model included no gravitational loading and simplified the mounting as a fixed
restraint at the transmission end of the axle and a axial/radial bearing restraint at the other. A
cumulative 96,000 foot pound torque was applied in 19,200 lb increments to each of the
Page 38 of 51
DE-EE0004573
FY2011
mounting bosses. This model had limited accuracy since the axial/radial bearing artificially
forced alignment, and no gravitational force over the axle span was accounted for. Nonetheless,
the model showed stress of 22,300 psi.
Figure 24: FEA results of axle torque test
The axle was designed to operate in fully reversed bending. Aluminum 6061 T-6 alloys have
fully reversed bending yield strength of 14,000 psi at 500 million cycles. This gives an
unacceptable factor of safety (0.627).
Eventually, the axle method of wheel support and torque transfer was abandoned because the
96000 foot pound torque required an exceptionally heavy axle, which incurred high material and
logistic costs. Additionally, the need for a more flexible application implied the requirement of a
modular design which could be applied in varying conditions.
The following requirements were determined for a satisfactory turbine section design:
1. High moment of inertia in torsion and transverse bending

2. Low weight
3. Ease of assembly
4. Modularity
A turbine comprised entirely of plastic was designed. However, although low stresses could be
maintained, it was not able to hold geometric stability and could not be applied.
Page 39 of 51
DE-EE0004573
FY2011
Figure 25: All plastic turbine design
Eventually a novel design using tubes mounted in an offset pattern from the center of end plates
of a modular turbine section was presented. Using a plurality of tubes further from the central
axis of rotation reduces the amount of material required for a desired polar or area moment of
inertia. The basic dimensions were analytically derived to achieve acceptable multi-axial stress
levels in bending and torque:
2
 x  y 
 max   x   y      xy2
 2 
where σ is the principal normal stress, and τ is the principal shear.
Page 40 of 51
DE-EE0004573
FY2011
Figure 26: Final turbine design
The analytical results led to more specific parameters for an FEA analysis using 12 inch tubes at
a five foot offset from center, and a thickness of 0.25 inches (see Figure 22). The FEA model
was evaluated using the Autodesk Inventor static analysis package. The built-in mesh generator
used 0.1" tetrahedral elements to produce a very realistic geometric mesh. It was desired to
refine the design and validate analytical predictions by modeling the combined stress of
transverse loading due to turbine weight and moment loading from turbine torque (the
anticipated max torque being operational torque plus log strike). To this end, a fixed constraint
was applied to the transmission flange, and a bearing constraint applied to the mounting flange
on the other side. The bearing constraint prevented radial motion but not axial or tangential
(rotational) motion. A torque was applied to the mounting flange (a limiting case whereby
torques are not evenly distributed along turbine sections, but concentrated at one end - not
anticipated during operation). A second load, gravitational acceleration, was employed along the
transverse axis to simulate turbine weight.
Page 41 of 51
DE-EE0004573
FY2011
Figure 27: FEA results, turbine frame torque test
The limitations of this FEA analysis were primarily two fold; first the bearing constraints in
Autodesk Inventor static analysis package could not specify a self-aligning bearing, giving the
turbine a false degree of angular restraint. Additionally, the effects of fatigue loading in fully
reversed bending could not be immediately evaluated using this software package.
However, it was concluded that the stress plots provided useful data insomuch that stresses were
generally low in the structure (under 5000 psi, as predicted by analytical methods).
Materials selection for the turbine section structure centered around acquiring a material which
possessed sufficient mechanical properties in terms of strength to weight ratio, fatigue and
corrosion resistance, and which could be easily machined, welded, and manufactured to
specification. Steel alloys were initially considered, due to their high fatigue resistance,
reasonable weldability and machinability, and comparatively low cost. However, steel alloys
typically require additional corrosion protection in marine environments, and are fairly heavy.
Various aluminum alloys were considered. 7075-T-651 aluminum features unimpressive
corrosion resistance, but has a 74,000 to 78,000 psi yield strength, and 23,000 psi fatigue
Page 42 of 51
DE-EE0004573
FY2011
strength in fully reversed bending at 500 million cycles9. Nevertheless it cannot be welded, and
is a high cost alloy. 6061-T6 aluminum has a 14,000 psi fatigue strength at 500 million cycles in
fully reversed bending10. 5086-H116 aluminum is a marine grade type aluminum with excellent
yield (30,000 psi) and fatigue strength (21,800 psi at 500,000,000 cycles fully reversed bending
stress)11 and light weight. Welding causes a local reduction in strength to O temper (yield
strength at 17,000 psi), but it can be welded, and a FEA analysis of stress locations predicted
acceptably low stress at welds. The excellent fatigue and corrosion resistance of 5086 series
aluminum alloys made it a preferred material for turbine and craft components.
Bearings and Adjustments
The turbine was specified with several inter-related design considerations in terms of mounting
and adjustment. First it was desirable for the transmission input shaft to act as the mounting
component for one side of the wheel. Second, the turbine should be vertically adjustable. That
is, it could be lowered into water for operation (and raised out for maintenance) without moving
craft. Finally, the structural potential for misalignment due to deck twisting moments and
vertical adjustment necessitated a robust self-aligning bearing system.
The design considerations concerning the transmission in particular will be discussed in more
detail in a later section; however it was specified with a low speed input flange rated for the
shear load of the turbine as well as the twisting moment. On the other side, a pillow block with
an integral self-aligning housing and precision plane bearing was specified. Both the pillow
block support and transmission assembly were designed to be fastened to vertically sliding
mounts actuated by linear actuation system. Initially, servo controlled linear actuators such as
those employed for machine tool positioning were discussed. These were abandoned in favor of
a more robust system, hand crank actuated screw jacks. Some concern about screw jacks
potentially failing under buckling was discussed; however the rating of the screw jacks specified
was considerably greater than the specified load required, and no risk of buckling was
anticipated.
9
Aerospace Specification metals Inc. Aluminum 7075 T-6, T-
651http://asm.matweb.com/search/SpecificMaterial.asp?bassnum=MA7075T6 Accessed Online, August 18, 2011
10
Aerospace Specification metals Inc. Aluminum 6061 T-6, T-651
http://asm.matweb.com/search/SpecificMaterial.asp?bassnum=MA6061t6 Accessed Online, August 18, 2011
11
Aerospace Specification metals Inc. Aluminum
http://asm.matweb.com/search/SpecificMaterial.asp?bassnum=MA5086H116 Accessed Online, August 18, 2011
Page 43 of 51
DE-EE0004573
FY2011
Figure 28: Assembly view- transmission and power generation components
Transmission
The general paradigm of the Poncelet turbine system is to provide low speed, high torque power
output. Optimum efficiency is obtained by running the wheel such that blade tip speed is 40% of
current speed, so a 15 foot diameter turbine constrained to run at optimum speed* in a 15 foot
per second current will rotate at approximately 8 revolutions per minute. Efficient electrical
power generation typically requires generator input with higher speeds and lower torques; to this
end, a transmission system was specified.
Several design requirements were formulated to narrow the field of potential transmissions. The
generator chosen was a low speed 36 pole AC permanent magnet generator with optimum
efficiency in the 150-200 RPM range (More discussion on this design paradigm will be
conducted in a later section). Thus the transmission must have a speed ratio of 30:1 with a 5-8
RPM input. The transmission must be weight and cost effective, and be readily mounted and
coupled to turbine and generator. It must be sufficiently compact to fit on the slider mechanism
and be sealed and protected from wind, silt and water. If lubricated, the lubricant must be
sufficiently sealed so as to present no environmental hazard.
Several transmission variants were considered for potential application. A caged belt drive was
initially considered. However, belt drives are most efficient at high speeds, and a 30:1 reduction
would require multiple sets of prohibitively large sheave/belt combinations. A chain drive would
be more efficient at low speeds, but would be large and require lubrication; additionally concerns
about noise pollution were discussed.
Page 44 of 51
DE-EE0004573
FY2011
It was concluded that most custom transmission solutions were excessively large and expensive;
therefore a commercial off the shelf option was explored. A Brevini epicyclic two-stage
transmission was selected. This transmission used planetary gear sets in series to provide
compact, light, zero maintenance transmission solution. This planetary gearset package offered
custom couplings and mounts on both high and low speed ends.
It was concluded that the transmission itself would be specified with tapped holes by which it
would be fastened to the sliding mount, which would bear the torque load of the generator
resistance. A love-joy gear coupling would provide misalignment tolerance on the turbine side,
and the generator would be rigidly mounted to the output side of the transmission.
Braking and Turbine Control
It was desired to maintain control over turbine speeds for three reasons: First in the event of
emergency, it would be desirable to stop the turbine. Secondly, concerns were discussed about
the risk of turbine "runaway" if inductive motor power output was exceeded. Finally, concerns
regarding inductive motor cut-in speeds, and the need to potentially slow and control wheel
rotation were discussed. More discussion regarding generators and generator controls is
available in a later section.
Initially, the general design paradigm regarding turbine design was as follows: the turbine would
provide motive power to the inductive generator, and any modification of rotational speed would
be executed mechanically. A hydraulic brake was specified to slow or stop the wheel if
necessary. This brake would require servo-actuation interfacing with the electrical generator
controls, or an operator to set cut-in speed and slow turbine if need be.
Due to design paradigm changes (discussed in more detail in a later section) regarding the
generator and control setup, a partnership with Energetic Drives led to the specification of a new
control system. This novel power generation/conditioning and controls system utilized a
permanent magnet generator, and automatically controlled generator resistance to provide
optimum torque and could cut in and out either automatically (by preset setpoints), manually (by
on-craft actuation) or remotely (by SCADA control).
The system designed by Energetic Drives allowed for considerable simplification of the
mechanical system, and elimination of numerous components, including the mechanical braking
system. In the current embodiment, the generator could be signaled by the control system to
provide back emf, or regenerative braking, and could cut in or out at any velocity that was
anticipated. This eliminated the hydraulic reservoir, lines, brakes, and actuation components.
Power Generation/ Conditioning

Specific application and design specifications were developed for pairing a generator with the
turbine prime mover. It was desired to employ a generator compatible with three modes of
application. The generator must function as a stand-alone power source (such as a backup power
source in the event of grid power failure); it must be capable of pairing with diesels to provide
Page 45 of 51
DE-EE0004573
FY2011
power, and it must function as a source connected to infinite grid power. Functionally, a
generator must be compact, inexpensive, robust, and tolerant of variance in rotor speed. To this
end, both permanent magnet and induction generator solutions were considered.
Induction generators function by forcing the prime mover to drive the rotor above a synchronous
speed, which is defined by following relation:
Hertz  RPS  Numberpairs
That is, the frequency of power generated depends on the rotational speed and number of pairs of
poles on the stator. Induction generators are typically larger, but lighter, for a given rated power
output than permanent magnet generators because they require no brushes or commutator. This
also makes them more rugged. Induction generators also tend to be less expensive than their
permanent magnet counterparts. The drawback of the induction generator is that they require a
source of excitation current for magnetizing flux; thus an induction generator is not a suitable
solution for stand-alone power. Additional concerns were that an error in controls could allow
an induction generator connected to an infinite grid to run at lower speeds (ie: the rotor turning
slower than rotating flux) whereby the machine would function like an induction motor and use
grid power to bring wheel up to speed.
A permanent magnet solution was selected, because although such generators are more
expensive, and contain more moving parts, they are more readily suitable for stand-alone power
production. An off the shelf solution was provided, which had 36 poles (enabling low speed
operation), and a wide operating band. Paired with an effective power signal conditioning
system (A more detailed discussion of the particulars of this system will be discussed later), this
solution offers a very acceptable range of operational productivity and flexibility.
Figure 29: Circuit diagram of Energetic Drives Active Front End system
Page 46 of 51
DE-EE0004573
FY2011
Electrics/ Controls/ Monitoring

Generator Control Algorithm and SCADA Controls
The operational imperatives driving craft design require a robust and complex feedback control
and monitoring system with integral remote access and supervision capacity. To this end, a
SCADA (supervisory control and data acquisition) system was specified. This system was
subject to several design requirements. First, it must modulate generator resistance to maintain
optimum ratio of blade tip speed to water speed, maintain cooperative master/slave power
sharing in the event of diesel pairing power production, and provide real-time data for water
velocity, wheel speed, voltage and reactive power production. Additionally, the SCADA system
must be capable of producing alarm outputs to an integral personnel alarm system in the event of
specific operating conditions.
Figure 30: Energetic Drives control cabinet
Generator Efficiency Optimization
The theoretical Poncelet efficiency was optimized by adhering to specific geometric and relative
velocity constraints. It was considered important for the control system to be designed such that
it would increase or decrease generator torque to maintain proper velocity for maximum power
output.
Page 47 of 51
DE-EE0004573
FY2011
Figure 31: Control cabinet components
Diesel Generator Pairing
As mentioned previously, it was desired to design the hydrokinetic turbine for three modes of
application. It had to be able to provide stand-alone power to a small grid; it had to be able to
pair with other small power sources (such as existing diesel generators) to power a grid, or it had
to be able to feed the infinite grid.
The standalone and infinite grid modes are fairly simple for a permanent magnet generator and
controls. However diesel pairing introduces a potentially problematic feedback loop since the
craft and most generators would have individual load sensing governor controls. Some concern
was discussed that this might result in an unstable response, which would not only affect power
output, but would cause oscillations which would cause diesel prime movers to run at inefficient
speeds.
To attenuate this concern, a PLC driven control system with Schweitzer relay sensors was
designed. If the relay sensed the activation/deactivation of another power source on a finite grid,
it signaled a master/slave set point control which set an optimal speed for the diesel prime
mover, and assigned the remaining load to the hydro-kinetic turbine.
Real-Time Monitoring and Control
Since the hydrokinetic turbine is designed for remote locations, it was considered desirable to
enable remote monitoring. Thus voltage output, river current speed, and wheel rotational speed
Page 48 of 51
DE-EE0004573
FY2011
would be available at a power plant computer remotely. Additionally, manual control and
programming changes to PLC set points could be executed remotely from a computer using the
SCADA interface.
Alarm Systems
Since the hydrokinetic turbine was designed for potential swift water deployment in remote
locations, some concerns were discussed concerning potential mooring component failure. A
positional monitoring system employing a Dynamic Global Positioning System coupled with an
excursion monitoring/reporting software package was specified for integration into the SCADA
control system. If the system sensed the craft moving outside of the defined excursion envelope,
an alarm would sound to indicate mooring cable failure; this system queries onboard GPS
sensors for craft position every five seconds, updates a five-year data-logged history of craft
positions and headings at a one-minute sampling rate, and additionally records alarms and events
in a data log.
The proposed positional monitoring system is tolerant of power outages and currently supports
the following industry standard communication protocols:
1. MODBUS RTU Over TCP

2. MODBUS ASCII/RTU/TCP
3. NMEA 0183
Means of Alerting Technicians
The proposed SCADA system interfaces with a Protalk CV3 alarm dialing system with cellular
amplification, integrated cellular module with voice and SMS text capabilities. This alarm
system is tolerant of power outages, and may be programmed for four different shifts, is highly
modular, and has low footprint. It will continue to dial numbers in its database until technicians
give confirmation of alarm notification.
The proposed system also has built-in radio port and public address systems which may be
programmed with redundant alert capability in after-hours situations.
An additional consideration for the SCADA monitoring/alarm system was alarm cascade. Since
the Protalk interface was capable of supporting a wide array of specific alarm messages from
digital and analog inputs, it was important that the acquisition and broadcast of craft data be
configured to give technicians optimum awareness of the mode of failure and extent in the event
of emergency involving several alarms from multiple component failures. The integrated PLC
interface would then organize the alarm cascade such that technicians would be able to
differentiate a transmission rotation stoppage caused by a debris jam from one caused by
mooring cable failure or transmission component failure. This allows emergency personnel and
technicians to best prepare themselves to address emergency situations.
Page 49 of 51
DE-EE0004573
FY2011
Partnerships and Collaboration
Energy Stability Solutions and Energetic Drives, LLC
During a survey of available power generation and control solutions, Hasz Consulting
collaborated with Energetic Drives, a company which provides power generation solutions
specifically for small (10 kW to 1.5 MW) projects. The "active front end" power generation
platform designed by Energetic Drives provides an exceptionally efficient production of clean
power, and significantly simplifies mechanical requirements in terms of turbine control.
Most engineered solutions for a variable speed prime mover involve a diode bridge rectifier to
convert "dirty" AC power to DC, and can only provide between 1-3% reactive power (volt-amp-
reactive) to compensate for reactive loads. The design utilized by Energetic Drives employs an
active bridge and LCL filter to eliminate harmonics while also providing the necessary reactive
power to maintain a stable power factor regardless of the phase of the load.
Potential Power Generation for Alaska and the Nation
A large portion of the project was completed by CE2 Engineers, Inc. of Anchorage, AK. As a
company with a successful history of completing highly technical projects in remote
communities, it was decided they would be an excellent partner to help determine the viability of
the chosen design. To this end, Hasz Consulting contacted them and they agreed to complete the
portions of the project which would analyze extractable energy estimates as well as the obstacles
to successful integration of the engineered system into the open market place.
Their full report is attached to this document. Their study was focused primarily on Alaskan
communities. These communities specifically, being remote, operating in harsh environmental
conditions and experiencing the highest cost of energy in the nation, are not only in the greatest
need, they are also the best test beds of this technology. The full study includes an estimate of the
extractable energy nationwide as well as the significant obstacles to full integration in the
market.
Conclusion
The project was considered successful in three ways. First, site specific engineering solutions
were developed for applying RISEC technology effectively in Alaskan river environment.
Secondly, the wide integration of stock components reduced design, construction and component
replacement costs. Finally, the modular construction design and flexible control system designed
by Energetic Drives, contributed toward methodology with broad potential application to remote
village power needs.
The effective deployment of RISEC technology in Alaskan rivers is tremendously significant in

light of the extremely high costs of energy in remote communities. Providing an energy source
which is both sustainable and economically feasible is crucial in preventing the extinction of
Page 50 of 51
DE-EE0004573
FY2011
remote Alaskan villages, many of which cannot remain financially solvent in the face of
increasing fossil fuel costs.
A significant aspect of this project was a study conducted by CE2 Engineers which compiled a
list of sites for potential commercial application of the project device. The study, which took
into account topography, river speed, and local community power needs, concluded that this
device would potentially be applicable at 46 village sites in Alaska alone, and 150 sites in the
continental United States.
RISEC implementation offers the benefit of flexible, comparatively inexpensive power solutions
for remote communities as well as providing a substantial step forward in the technical and
commercial viability of alternative technology.
Page 51 of 51
Riverine Resource Assessment
Poncelet Kinetics RHK100 Hydrokinetic Device
Prepared by:
CE2 Engineers, Inc.
DRAFT
Report Date:
February 2011
Communities located on rivers in Alaska may potentially benefit by integrating an RHK100 hydrokinetic
device into their existing power system. Potential benefit will depend on the community’s proximity to
water, the estimated stream velocity, the distance between the river and the community, the amount of
summer power supplanted, and the future changes in community population and fuel costs. Benefits
result from displacing some or all of a community’s summer power load. Benefits to larger communities
like Mountain Village, Tok, and Galena, will differ from those to smaller communities like Red Devil,
Sleetmute, and McGrath.
PONCELET KINETICS RHK100 RIVERINE RESOURCE ASSESSMENT
Contents
I. Introduction ............................................................................................................................. 1
II. Method .................................................................................................................................... 2
A. Community List ................................................................................................................................. 2
B. River Velocity Data ............................................................................................................................ 3
C. Community Distance to River ........................................................................................................... 3
D. Suitable River Hydrokinetic Power Capacity ..................................................................................... 3
E. Potential Monetary Benefits of the RHK100..................................................................................... 4
III. Results and Discussion ............................................................................................................. 6
A. Community List ................................................................................................................................. 6
B. River Velocity Data ............................................................................................................................ 6
C. Community Distance to River ........................................................................................................... 7
D. Suitable River Hydrokinetic Power Capacity ..................................................................................... 7
E. Potential Monetary Benefits of the RHK100..................................................................................... 8
List of Appendices
Appendix A—Alaska Community List
Appendix B—Analysis of Alaska Community Power Consumption and Potential Energy Offset
CE2 Engineers, Inc. i Report Date: February 2011

I. Introduction
Whitestone Power and Communications (WPC) is in the process of developing and trademarking an
electrical generation device called “the Poncelet Kinetics RHK100”. This River In-Stream Energy
Conversion (RISEC) device, can be used in communities located in proximity to a sufficient water
resource to generate electrical power from the hydrokinetic water flow. The RHK100 is a pontoon-
mounted Poncelet undershot water wheel (estimated at 12-foot-wide with a 16-foot diameter) with a
nominal electrical power output capacity of 100 kW. The float footprint is estimated at 34 feet by 19
feet, with a weight of approximately 15,000 pounds. The installation will be moored to the shore and
protected with Coast Guard-approved safety equipment.
WPC contracted CE2 Engineers, Inc. (CE2) to perform a preliminary assessment of Alaska communities
located near flowing rivers that might benefit from integrating a hydrokinetic device into its present
power generation system. The assessment includes:
1. a list of communities situated near Alaska rivers where the RHK100 would be suitable,
2. an inventory and summary of existing river velocity data for rivers near these communities,
3. a description of the approximate distance between the river and the community’s power
plant,
4. an estimation of the potential amount of hydrokinetic-derived power needed for each
community, and
5. an estimation of the potential monetary benefit to each community from the maximum
suitably-sized RHK100 hydrokinetic device.
This assessment assumes that the production capacity of the RHK100 can be adjusted to work in rivers
with velocity/flow rates as low as 1 foot/second, and depths of three feet or greater.
CE2 Engineers, Inc. 1 Report Date: February 2011

II. Method
The five pieces of the assessment were done mainly independent of one another. The methods
employed in each piece of the assessment are described here.
A. Community List
A community list was developed by consulting the data published on the website of the Alaska
Community Database maintained by the State of Alaska, Department of Commerce, Community,
and Economic Development (DCCED). From a list of all currently identified Alaska communities
(presented in Appendix A), the following communities were initially eliminated:
Communities listed with zero or very small population,

Communities with no electrical power distribution system,
North Slope communities (due to the extremely short ice-free season for
the Arctic rivers)
Communities whose geography includes the absence of viable rivers, including no rivers,
extremely slow-moving rivers, steep rivers, and rivers subject to the influence of tides were not
evaluated, nor were those communities in the Railbelt. The list was further reduced to those
located on major river systems, away from the lower, slower reaches of the rivers (such as
Yukon River communities like Alakanuk and Kotlik, or Kuskokwim River communities like
Tuluksak).
The result was a list of 45 communities primarily situated on major Alaska Rivers. From north to
south, generally, those rivers include: Noatak, Kobuk, Koyukuk, Yukon, Tanana, Kuskokwim,
Kvichak, Nabesna, and Copper. The following matrix is a summary list of the communities
initially selected for the RHK100 device assessment.
River Community
Copper Chitina, Slana
Kobuk Ambler, Kiana, Shungnak
Koyukuk Allakaket/Alatna, Bettles, Hughes, Huslia
Kuskokwim Aniak, Chuathbaluk, Crooked Creek, Kalskag (Upper and Lower), McGrath,
Nikolai, Red Devil, Sleetmute, Stony River
Kvichak Igiugig
Noatak Noatak
Nabesna Northway
Tanana Tok/Dot Lake/Tanacross, Tanana, Manley Hot Springs

River Community
Yukon Anvik, Circle, Fort Yukon, Galena, Grayling, Holy Cross, Kaltag, Koyukuk,
Marshall, Mountain Village, Nulato, Pilot Station, Ruby, Russian Mission, St.
Mary’s, Pitka’s Point, Steven’s Village
B. River Velocity Data
Data that describes flow rates for Alaska rivers and streams exists in several places on the web:
the United States Geological Survey (USGS) National Water Information System at
http://wdr.water.usgs.gov/nwisgmap/; the State of Alaska Department of Natural Resources
(DNR), Alaska Hydrologic Survey Streams Database at
http://dnr.alaska.gov/mlw/water/hydro/streams.cfm; and a presentation entitled “Assessment
of Hydrokinetic Energy Resources in Alaska Rivers” written by faculty and staff at the University
of Alaska Anchorage, Department of Engineering, and published on the Alaska Energy Authority
website at http://www.akenergyauthority.org/OceanRiver/TomRavens_REC4-2010.pdf.
Research staff at CE2 were surprised by the apparent scarcity of stream velocity data. However,
several knowledgeable hydrologists with the USGS, National Park Service, and the Bureau of
Land Management confirmed that stream velocity data are very rare in Alaska, mainly due to
cost of recording and collecting those data.
C. Community Distance to River
Google Earth mapping tools were used to approximate the distance between the river channel
and the vicinity of the power plant. Distances were rounded up to the nearest 100 feet.
D. Suitable River Hydrokinetic Power Capacity
The amount of river hydrokinetic power suitable for use in a community’s existing power system
will depend on the average load (measured in kilowatts) carried by that community during the
months the RHK100 device would be working. The State of Alaska, Alaska Energy Authority’s
Power Cost Equalization (PCE) program published data on the power production, fuel and non-
fuel costs, and population for many communities in rural Alaska.
For those communities initially deemed viable (see II.A), PCE data were collected, when
available. Not all viable communities were included on the PCE roster, and some were on the
roster but contained limited power production data. Some communities share power via an
intertie, such as Tok, Tanacross, and Dot Lake, and similar groupings were treated as one entity.

Because the RHK100 device will only operate during summer months (estimated here to be the
four-month period from May 15 to September 15), an estimate of the average summer load was
necessary to estimate how many kilowatts would be available to offset. Although a few
communities have facilities such as fish processing plants that create a higher load on their
community power system during the summer, most of these examples are balanced out by the
power demands of the school building during the non-summer months.
Brent Petrie, Manager of Community Development at the Alaska Village Electric Cooperative
(AVEC), which owns and operates power production and distribution systems in some 50 rural
communities, estimated that the average community will require ten percent less power during
the summer months than they will during the non-summer months. That is the measure used to
determine the average monthly summer load for each community with published PCE data.
The Analysis of Alaska Community Power Consumption and Potential Energy Offset, presented
in Appendix B, contains most of the data discussed in this assessment. As shown in this table,
some communities have a summer load as low as 13 kilowatts (Stony River) or 15 (Red Devil),
while others have loads of 300 kilowatts (St. Mary’s/Pitka’s Point, and Fort Yukon) and 400
kilowatts per month (Galena). The Tok/Tanacross/Dot Lake intertie pulls a load of approximately
1,200 kilowatts during an average summer month.
It is important to note that the published PCE data covers community power information for the
years 2002-2009. A statistical method of projecting future numbers based on a known trend,
“least square”, was used to project power production, cost, and population data through the
year 2020. An “average year”, here and elsewhere in this report, represents the average
PROJECTED figure for the twelve years from 2009-2020.
E. Potential Monetary Benefits of the RHK100
The monetary benefit of using an RHK100 hydrokinetic device will vary between communities
insofar as each community has a unique set of energy production and cost characteristics. While
the amount of diesel-generated electricity may be similar between communities (and,
correspondingly, the amount of diesel displaced) the cost of generating that electricity will vary
between any two communities. For the community where diesel-produced energy costs more,
supplemental power from the RHK100 will realize greater cost savings than for a community
where energy is more affordable. The customer base and distribution of costs will also affect the
potential benefit of supplementing a community’s electrical production, where small
communities pay more per capita than their more populous neighbors.

“Least square” trend lines were produced for power production, fuel costs, and population for
each community. These projections varied greatly from one community to the next, being based
on the fluctuations over the years 2002-2009, the years for which PCE data exist.
These data were used to make a general determination of: 1) the cost of producing power using
diesel for each community; and 2) the amount of cost savings resulting from the RISEC-derived
supplementary power initially identified as suitable for each community. Monetary benefit was
calculated to reflect both the average annual per-capita cost savings, as well as the average
annual overall cost savings to the community.
The amounts presented in this study for cost savings do not account for any of the costs for
RHK100 construction (capital costs) or maintenance. And while the community’s average non-
fuel related expenses are shown in Appendix B, they do not affect any of the cost savings
figures; cost savings only reflect displacement of diesel.

III. Results and Discussion

A. Community List
The matrix introduced on page 2 of this report contains the list of Alaska communities initially
judged viable for introduction of the RHK100 hydrokinetic device. These communities are, using
the best available information: situated close to a river, where the flow rate is sufficiently rapid,
where the river and the channel is not too steep or too shallow, and where the influence of
coastal tides, waves, and shifts in flow direction will not interfere with the RHK100’s operation.
Other communities are listed as non-viable due to one of the following reasons: insufficient
population base, coastal influence, insufficient water flow, intertie-connected. Several locations
were rejected because they are government facilities, such as Eielson AFB, and two were
rejected as corporate or private utilities. Certain communities are marked non-viable for
“environmental” considerations; where the presence of a popular sport fishery or an urban
setting was seen as a significant impediment to the RHK100 hydrokinetic device
implementation.
Twenty seven additional communities are considered potential sites for implementation of a
hydrokinetic device, but additional data must be gathered for a full assessment. These
communities are marked in Appendix A with the selection code “P”.
B. River Velocity Data
Appendix B presents the summary data discussed in this report, including river velocity
measurements from the three sources mentioned above.
River velocity data McGrath, Aniak, and Sleetmute were extracted from the DNR Hydrologic
Survey database, found at http://dnr.alaska.gov/mlw/water/hydro/streams.cfm.
Stream data for Kalskag, Aniak, Chuathbaluk, Mountain Village, Saint Mary’s, Pilot Station,
Marshall, Holy Cross, Anvik, Grayling, Nulato, Koyukuk, and Galena were drawn from UAA
Professor Tom Ravens’ presentation prepared for AEA and titled “Assessment of Hydrokinetic
Energy Resources in Alaska Rivers, which is found at
http://www.akenergyauthority.org/OceanRiver/TomRavens_REC4-2010.pdf.
Stream data for Chitina, Kiana, Noatak, Tanana, Steven’s Village, and Stony River were extracted
from the USGS database, found at http://wdr.water.usgs.gov/nwisgmap/.

For most an average flow rate, in feet per second, as well as a maximum flow rate are shown in
the Analysis of Community Power Consumption and Potential Energy Offset presented in
Appendix B. The measurements of river velocity vary widely, even for the same community. For
example, flow rates recorded for the Noatak River near the village of Noatak average 0.9
feet/second, but a maximum rate was reported on the same day as 4.49 feet/second. Because
the RHK100 will produce power at an even, fixed level of output, the average velocity figure
should be considered the baseline for devising the mechanical transfer of water flow to
electricity production.
Both the USGS and the DNR websites have large amounts of stream data, but most of that is for
streams that are near villages but are nonetheless very small, and slow-moving. These were not
included in this analysis.
C. Community Distance to River
Appendix B identifies the approximate distances between each community and the river where
the RHK100 would be placed. Many communities are located very close to their rivers, and 500
feet was the standard estimated distance incremental value. Some communities are farther
away, between 1,000 feet and 4,000 feet. For Manley Hot Springs, a distance of approximately
two (2) miles separates the community from the Tanana River channel.
A shorter distance would certainly result in a less-expensive, more efficient connection of the
RHK100 to the community power grid, and a greater distance would be correspondingly more
expensive and less efficient. The resulting expense and efficiency, however, are not included in
this analysis.
D. Suitable River Hydrokinetic Power Capacity
The Analysis of Alaska Community Power Consumption and Potential Energy Offset (Appendix B)
includes a column (“Water Turbine kW”) that assigns a number of kilowatts to each community.
This number represents a river hydrokinetic energy output level, in kilowatts, suggested for each
community as a supplement to its summer energy load. In some cases, this amount will supplant
a good portion of the community’s summer energy load. In others, it will provide for the entire
load amount.
Because Red Devil and Stony River have such a low summer load level, an RHK100 hydrokinetic
device that produces 25 kilowatts will result in maximum savings for those communities. For the
St. Mary’s/Pitka’s Point as well as for the Tok/Tanacross/Dot Lake intertie, an RHK100 producing
400 kilowatts appears optimal. Because WPC’s initial RHK100 hydrokinetic device proposal

indicated a per-device output of 25 kilowatts, the figures identified in this column are in
increments of 25kW.
The relationship between summer load levels, stream velocity, and the RHK100 capacity was
only cursorily addressed in discussions between CE2 and WPC. WPC indicated that the RHK100
could be adjusted down for a low load situation, as well as ramped up at will to supplant larger
loads. Therefore, nearly all communities show the RHK100 supplanting 100% of the power
production requirements for the four-month-long summer period. Mathematically, that would
eliminate the need to produce diesel power for one-third of the year (resulting in a 30% annual
reduction, which accounts for the slightly higher power usage during the non-summer months).
E. Potential Monetary Benefits of the RHK100
Appendix B shows the “Yearly Production Cost Savings” for the RHK100 device configuration
recommended for each community. Based on the projections tracking the cost of fuel (an
increase, in nearly all cases) and the amount of power required (increase in some cases,
decrease in others), supplementing summer power requirements with RHK100 devices will
produce annual overall community cost savings in the $20,000 range for communities like
Koyukuk; the $30,000 range for communities like Ruby, Chitina, and Hughes); upwards to annual
savings of $300,000 and $400,000 in places like McGrath, Galena, St. Mary’s/Pitka’s Point, and
Fort Yukon.
The annual per-capita cost savings measure shows how the members of a smaller community
may benefit, since fewer people will pay a greater percentage of the community’s total power
costs and will benefit more from an overall reduction in community power production costs.
Red Devil, for example, may only save $27,177 yearly with a 25kW RHK100 device, but based on
the high per-capita cost of energy in that community, each individual may save an estimated
$1,182 annually.
Figure 1, on the next page, illustrates both the annual per capita savings and the annual overall
community savings for all of the communities evaluated.
Because the analysis presented here does not account for any of the costs for RHK100 device
construction or maintenance, these cost savings figures must only be viewed as a baseline from
which to develop a more extensive, precise economic assessment.

Figure 1
Annual Projected Community Cost Savings: Per Capita and Total

$1,600 $500,000
$450,000
$1,400
$400,000
$1,200
$350,000
$1,000
$300,000
$800 $250,000
$200,000
$600
$150,000
$400
$100,000
$200
$50,000
$0 $0

Appendix A
Alaska Community List
from State of Alaska Department of Commerce, Community, and Economic Development Community Database
DCCED Evaluation
COMMUNITY POPULATION Code Notes
Adak 165 NE Tidal Influence
Afognak 0 R Insufficient Population
Akhiok 80 NE Water flows require further evaluation
Akiachak 645 NE Water flows need further evaluation
Akiak 346 NE Water flows need further evaluation
Akutan 846 NE Water flows need further evaluation
Alakanuk 686 NE Water flows need further evaluation
Alatna 22 E
Alcan Border 26 NE Water flows need further evaluation
Aleknagik 229 P
Aleneva 67 R No Power Distrib System
Allakaket 100 E
Alpine 0 R Insufficient Population
Ambler 261 E
Anaktuvuk Pass 287 NE Water flows need further evaluation
Anchor Point 1772 NE Railbelt
Anchorage 290588 NE URBAN
Anderson 275 NE Railbelt
Andreafsky 140 P
Angoon 442 NE Water flows need further evaluation
Aniak 485 E
Anvik 75 E
Arctic Village 139 NE Water flows need further evaluation
Atka 71 NE Hydro in use
Atmautluak 296 NE Water flows need further evaluation
Atqasuk 201 R Short Season
Attu Station 15 R Govt Facility
Barrow 4119 R Short Season
Bear Creek 2009 NE Water flows need further evaluation
Beaver 58 P
Belkofski 0 R Insufficient Population
Beluga 24 NE Railbelt
Bethel 5803 NE Water flows need further evaluation
Bettles 19 E
Big Delta 840 NE Railbelt
Big Lake 3331 NE Railbelt
Bill Moore's Slough 0 R Insufficient Population
Birch Creek 31 P
Brevig Mission 358 NE Water flows need further evaluation
Evaluation code:
E = Evaluated
NE or P = Potential sites, require additional data for evaluation
R = Rejected Page 1 of 10
Appendix A
DCCED Evaluation
Buckland 432 NE Water flows need further evaluation
Buffalo Soapstone 738 NE Railbelt
Butte 3255 NE Railbelt
Cantwell 200 NE Railbelt
Central 96 NE Water flows need further evaluation
Central 96 NE Water flows need further evaluation
Chalkyitsik 60 NE Water flows need further evaluation
Chase 35 R No Power Distrib System
Chefornak 475 NE Water flows need further evaluation
Chenega Bay 71 NE Water flows need further evaluation
Chevak 945 NE Water flows need further evaluation
Chickaloon 277 NE Railbelt
Chicken 23 R No Power Distrib System
Chignik Lagoon 73 NE Water flows need further evaluation
Chignik Lake 105 NE Water flows need further evaluation
Chignik 62 NE Water flows need further evaluation
Chiniak 48 NE Water flows need further evaluation
Chisana 9 R No Power Distrib System
Chistochina 95 P
Chitina 117 E
Chuathbaluk 111 E
Chuloonawick 0 R Insufficient Population
Circle 99 E
Clam Gulch 166 NE Water flows need further evaluation
Clark's Point 61 NE Water flows need further evaluation
Coffman Cove 192 NE Water flows need further evaluation
Cohoe 1332 NE Water flows need further evaluation
Cold Bay 84 P
Coldfoot 13 R No Power Distrib System
College 12552 NE Railbelt
Cooper Landing 344 NE Railbelt
Copper Center 297 P
Copperville 131 P
Cordova 2126 NE Water flows need further evaluation
Council 8 R No Power Distrib System
Covenant Life 89 P
Craig 1101 NE Water flows need further evaluation
Crooked Creek 131 E
Crown Point 77 NE Water flows need further evaluation
Evaluation code:
E = Evaluated
Appendix A
DCCED Evaluation
Cube Cove 0 R Insufficient Population
Deering 118 P
Delta Junction 1128 NE Railbelt
Deltana 2355 NE Railbelt
Diamond Ridge 860 P
Dillingham 2264 NE Water flows need further evaluation
Diomede 117 NE Water flows need further evaluation
Dot Lake Village 37 E (see Tok/Dot Lake/Tanacross)
Dot Lake 16 E (see Tok/Dot Lake/Tanacross)
Douglas 4890 NE Water flows need further evaluation
Dry Creek 87 R No Power Distrib System
30,000
Eagle River-Chugiak (2000 pop.) NE Railbelt
Eagle Village 54 P
Eagle 146 P
Edna Bay 49 R No Power Distrib System
Eek 282 NE Water flows need further evaluation
Egegik 73 NE Water flows need further evaluation
Eielson AFB 2896 R Govt Facility
Eklutna 384 NE Railbelt
Ekuk 0 R No Power Distrib System
Ekwok 109 NE Water flows need further evaluation
Elfin Cove 25 NE Water flows need further evaluation
Elim 337 NE Water flows need further evaluation
Emmonak 774 NE Water flows need further evaluation
Ester 2034 NE Railbelt
Evansville 13 P
Excursion Inlet 11 R No Power Distrib System
Eyak 107 NE Water flows need further evaluation
Fairbanks 32506 NE Railbelt
False Pass 41 NE Water flows need further evaluation
Farm Loop 1313 NE Railbelt
Ferry 36 R No Power Distrib System
Fishhook 3337 NE Railbelt
Flat 0 R No Power Distrib System
Fort Greely 413 R Govt Facility
Fort Yukon 585 E
Four Mile Road 39 NE Railbelt
Fox River 604 NE Water flows need further evaluation
Evaluation code:
E = Evaluated
Appendix A
DCCED Evaluation
Fox 390 NE Railbelt
Fritz Creek 1818 NE Water flows need further evaluation
Funny River 796 NE Water flows need further evaluation
Gakona 202 P
Galena 564 E
Gambell 666 NE Water flows need further evaluation
Game Creek 16 R No Power Distrib System
Gateway 4068 NE Railbelt
Georgetown 3 R No Power Distrib System
Girdwood 2000 NE Railbelt
Glacier View 246 NE Railbelt
Glennallen 473 NE Water flows need further evaluation
Golovin 154 NE Water flows need further evaluation
Goodnews Bay 237 NE Water flows need further evaluation
Grayling 168 E
Gulkana 131 P
Gustavus 451 NE Water flows need further evaluation
Haines Borough 2286 P
Halibut Cove 27 NE Water flows need further evaluation
Hamilton 0 R No Power Distrib System
Happy Valley 561 NE Water flows need further evaluation
Harding-Birch Lakes 287 NE Railbelt
Healy Lake 10 NE Water flows need further evaluation
Healy 1002 NE Railbelt
Hobart Bay 1 R No Power Distrib System
Hollis 193 NE Water flows need further evaluation
Holy Cross 187 E
Homer 5551 NE Water flows need further evaluation
Hoonah 764 NE Water flows need further evaluation
Hooper Bay 1158 NE Water flows need further evaluation
Hope 151 NE Railbelt
Houston 1664 NE Railbelt
Hughes 83 E
Huslia 265 E
Hydaburg 340 NE Water flows need further evaluation
Hyder 87 R Terrain
Igiugig 64 E
Iliamna 91 P
Ivanof Bay 0 R Insufficient Population
Evaluation code:
E = Evaluated
Appendix A
DCCED Evaluation
Jakolof Bay 0 R Insufficient Population
Juneau 30661 NE Hydro in use
Kachemak 430 R Environmental sensitivity
Kaguyak 0 R Insufficient Population
Kake 497 NE Water flows need further evaluation
Kaktovik 286 R Short Season
Kalifornsky 7495 R Environmental sensitivity
Kaltag 172 E
Kanatak 0 R No Power Distrib System
Karluk 38 NE Water flows need further evaluation
Kasaan 56 NE Water flows need further evaluation
Kasigluk 567 NE Water flows need further evaluation
Kasilof 536 R Environmental sensitivity
Kenai 7115 R Environmental sensitivity
Kenny Lake 412 NE Water flows need further evaluation
Ketchikan 7503 NE Water flows need further evaluation
Kiana 374 E
King Cove 744 P
King Island 0 R No Power Distrib System
King Salmon 383 NE Water flows need further evaluation
Kipnuk 671 NE Water flows need further evaluation
Kivalina 410 NE Water flows need further evaluation
Klawock 782 NE Water flows need further evaluation
Klukwan 72 P
Knik River 631 NE Railbelt
Knik-Fairview 13824 NE Railbelt
Kobuk 122 P
Kodiak Station 1321 R Govt Facility
Kodiak 6626 NE Water flows need further evaluation
Kokhanok 184 NE Water flows need further evaluation
Koliganek 182 NE Water flows need further evaluation
Kongiganak 465 NE Water flows need further evaluation
Kotlik 618 NE Water flows need further evaluation
Kotzebue 3154 NE Water flows need further evaluation
Koyuk 358 E
Koyukuk 105 E
Kupreanof 24 R No Power Distrib System
Kwethluk 764 NE Water flows need further evaluation
Kwigillingok 365 NE Water flows need further evaluation
Evaluation code:
E = Evaluated
Appendix A
DCCED Evaluation
Lake Louise 100 NE Railbelt
Lake Minchumina 17 NE Water flows need further evaluation
Larsen Bay 79 NE Water flows need further evaluation
Lazy Mountain 1446 NE Railbelt
Levelock 88 NE Water flows need further evaluation
Lime Village 19 P
Livengood 24 R No Power Distrib System
Lowell Point 76 NE Water flows need further evaluation
Lower Kalskag 251 E
Lutak 38 R No Power Distrib System
Manley Hot Springs 81 E
Manokotak 438 NE Water flows need further evaluation
Marshall 414 E
Mary's Igloo 0 R No Power Distrib System
McCarthy 51 R No Power Distrib System
McGrath 322 E
McKinley Park 168 NE Railbelt
Meadow Lakes 7319 NE Railbelt
Mekoryuk 174 NE Water flows need further evaluation
Mendeltna 57 NE Water flows need further evaluation
Mentasta Lake 120 NE Water flows need further evaluation
Metlakatla 1499 NE Water flows need further evaluation
Meyers Chuck 16 R No Power Distrib System
Miller Landing 0 R Insufficient Population
Minto 191 NE Water flows need further evaluation
Moose Creek 729 NE Railbelt
Moose Pass 189 NE Railbelt
Mosquito Lake 235 NE Water flows need further evaluation
Mountain Village 782 E
Mud Bay 178 R No Power Distrib System
Naknek 516 NE Water flows need further evaluation
Nanwalek 226 NE Water flows need further evaluation
Napaimute 0 R No Power Distrib System
Napakiak 337 NE Water flows need further evaluation
Napaskiak 428 NE Water flows need further evaluation
Naukati Bay 118 NE Water flows need further evaluation
Nelchina 51 NE Water flows need further evaluation
Nelson Lagoon 60 NE Water flows need further evaluation
Nenana 479 NE Railbelt
Evaluation code:
E = Evaluated
Appendix A
DCCED Evaluation
New Allakaket 37 P
New Stuyahok 519 NE Water flows need further evaluation
Newhalen 162 P
Newtok 355 NE Water flows need further evaluation
Nightmute 264 NE Water flows need further evaluation
Nikiski 4465 NE Water flows need further evaluation
Nikolaevsk 315 NE Water flows need further evaluation
Nikolai 87 E
Nikolski 33 NE Water flows need further evaluation
Ninilchik 824 R Environmental sensitivity
Noatak 486 E
Nome 3468 NE Water flows need further evaluation
Nondalton 186 NE Water flows need further evaluation
Noorvik 628 R Tidal Influence
North Pole 2200 NE Railbelt
Northway Junction 60 P
Northway Village 76 P
Northway 88 E
Nuiqsut 424 R Short Season
Nulato 240 E
Nunam Iqua 193 NE Water flows need further evaluation
Nunam Iqua 193 NE Water flows need further evaluation
Nunapitchuk 539 NE Water flows need further evaluation
Ohogamiut 0 R No Power Distrib System
Old Harbor 193 NE Water flows need further evaluation
Oscarville 109 NE Water flows need further evaluation
Ouzinkie 170 NE Water flows need further evaluation
Paimiut 2 R No Power Distrib System
Palmer 5532 NE Railbelt
Pauloff Harbor 0 R No Power Distrib System
Paxson 16 NE Private utility
Pedro Bay 48 NE Water flows need further evaluation
Pelican 122 R Terrain
Perryville 122 NE Water flows need further evaluation
Petersburg 2973 NE Water flows need further evaluation
Petersville 6 R No Power Distrib System
Pilot Point 66 NE Water flows need further evaluation
Pilot Station 577 E
Pitkas Point 113 E
Evaluation code:
E = Evaluated
Appendix A
DCCED Evaluation
Platinum 57 NE Water flows need further evaluation
Pleasant Valley 765 NE Railbelt
Point Baker 11 R No Power Distrib System
Point Hope 713 R Short Season
Point Lay 234 R Short Season
Point MacKenzie 273 NE Railbelt
Pope-Vannoy Landing 5 R No Power Distrib System
Port Alexander 61 R No Power Distrib System
Port Alsworth 118 NE Water flows need further evaluation
Port Clarence 23 R Govt Facility
Port Graham 137 NE Water flows need further evaluation
Port Heiden 83 NE Water flows need further evaluation
Port Lions 200 NE Water flows need further evaluation
Port Protection 72 R No Power Distrib System
Port William 0 R No Power Distrib System
Portage Creek 7 R No Power Distrib System
Primrose 65 NE Water flows need further evaluation
Prudhoe Bay 3 R Short Season
Prudhoe Bay 3 R Short Season
Quinhagak 680 NE Water flows need further evaluation
Rampart 12 R Insufficient Population
Red Devil 44 E
Red Dog Mine 35 NE Corporate generators
Ridgeway 2050 NE Water flows need further evaluation
Ruby 149 E
Russian Mission 363 E
Saint George 111 NE Water flows need further evaluation
Saint Mary's 553 E
Saint Michael 446 NE Water flows need further evaluation
Saint Paul 459 NE Water flows need further evaluation
Salamatof 855 R Environmental sensitivity
Salcha 985 NE Railbelt
Sand Point 1001 NE Water flows need further evaluation
Savoonga 721 NE Water flows need further evaluation
Saxman 434 NE Water flows need further evaluation
Scammon Bay 528 NE Water flows need further evaluation
Selawik 849 NE Water flows need further evaluation
Seldovia Village 166 NE Water flows need further evaluation
Seldovia 265 NE Water flows need further evaluation
Evaluation code:
E = Evaluated
Appendix A
DCCED Evaluation
Seward 2609 NE Water flows need further evaluation
Shageluk 97 NE Water flows need further evaluation
Shaktoolik 231 NE Water flows need further evaluation
Shemya Station 27 R Govt Facility
Shishmaref 606 NE Water flows need further evaluation
Shungnak 270 E
Silver Springs 198 P
Sitka 8627 R Terrain
Skagway 865 NE Water flows need further evaluation
Skwentna 73 R No Power Distrib System
Slana 102 E
Sleetmute 71 E
Soldotna 4021 R Environmental sensitivity
Solomon 0 R Insufficient Population
South Naknek 68 NE Water flows need further evaluation
Stebbins 605 NE Water flows need further evaluation
Sterling 5348 R Environmental sensitivity
Stevens Village 64 E
Stony River 48 E
Sunrise 19 NE Railbelt
Susitna 16 R No Power Distrib System
Sutton-Alpine 1407 NE Railbelt
Takotna 53 NE Water flows need further evaluation
Talkeetna 894 NE Railbelt
Tanacross 203 E (see Tok/Dot Lake/Tanacross)
Tanaina 7407 NE Railbelt
Tanana 251 E
Tatitlek 83 NE Water flows need further evaluation
Tazlina 207 NE Water flows need further evaluation
Telida 3 R Insufficient Population
Teller 261 NE Water flows need further evaluation
Tenakee Springs 104 R Terrain
Tetlin 169 NE Water flows need further evaluation
Thom's Place 6 R Insufficient Population
Thorne Bay 424 NE Water flows need further evaluation
Togiak 820 NE Water flows need further evaluation
Tok 1429 E
Toksook Bay 596 NE Water flows need further evaluation
Tolsona 26 NE Water flows need further evaluation
Evaluation code:
E = Evaluated
Appendix A
DCCED Evaluation
Tonsina 78 NE Water flows need further evaluation
Trapper Creek 444 NE Railbelt
Tuluksak 471 NE Water flows need further evaluation
Tuntutuliak 384 NE Water flows need further evaluation
Tununak 330 NE Water flows need further evaluation
Twin Hills 74 NE Water flows need further evaluation
Two Rivers 663 NE Railbelt
Tyonek 166 NE Railbelt
Uganik 0 R Insufficient Population
Ugashik 15 R No Power Distrib System
Umkumiute 0 R No Power Distrib System
Unalakleet 725 NE Water flows need further evaluation
Unalaska 3662 NE Water flows need further evaluation
Unga 0 R No Power Distrib System
Upper Kalskag 223 E
Valdez 4498 R Terrain
Venetie 185 NE Water flows need further evaluation
Wainwright 551 R Short Season
Wales 148 NE Water flows need further evaluation
Wasilla 7245 NE Railbelt
Whale Pass 60 NE Water flows need further evaluation
White Mountain 202 NE Water flows need further evaluation
Whitestone 173 NE Railbelt
Whittier 159 NE Railbelt
Willow Creek 157 NE Water flows need further evaluation
Willow 2218 NE Railbelt
Wiseman 16 R No Power Distrib System
Womens Bay 740 NE Water flows need further evaluation
Woody Island 0 R No Power Distrib System
Wrangell 2058 NE Water flows need further evaluation
Y 1057 NE Railbelt
Yakutat 628 NE Water flows need further evaluation
Total 636500 382
Evaluation code:
E = Evaluated
CE2 Engineers, Inc. Whitestone Power & Communications
Analysis of Alaska Community Power Consumption and Potential Energy Offset
Yearly Yearly
Avg Max Est intertie Avg Non-Fuel Average Annual Average Summer Load Load Summer Production/ Production/ Avg
Velocity Velocity distance Avg Fuel Cost Cost kWh (2009- Avg Ratio Non- Summer Load Range Range Water Production Cost Savings Cost Savings Population Per Capita Cost
Community River (Ft/Sec) (Ft/Sec) (ft) Per kWh (2009-2020) 2020) Fuel to Fuel (2009-2020) (2009-2020) Trend turbine kW Offset (Pct) ($) (2009-2020) Savings
Allakaket/Alatna Koyukuk 500 $0.46 $121,985 703,077 0.38 72 68-76 Up 100 100% 30% $96,646 132 $732
Ambler* Kobuk 700 $0.62 $350,750 1,385,110 0.41 142 127-148 Up 150 100% 30% $255,872 249 $1,028
Aniak Kuskokwim 3.70 9.00 500 $0.50 $1,042,858 2,541,020 0.83 261 266-262 Even 300 100% 30% $378,902 465 $815
Anvik* Yukon 3.70 9.00 1,750 $0.49 $123,050 445,992 0.56 46 41-45 Up 75 100% 30% $65,420 86 $761
Bettles Koyukuk 500 $0.45 $117,802 464,822 0.56 48 67-33 Down 100 100% 30% $63,373 58 $1,093
Chitina Copper 5.57 7.25 4,000 $0.32 $60,320 392,778 0.48 40 44-37 Down 50 100% 30% $38,042 111 $343
Chuathbaluk Kuskokwim 4.00 8.00 500 $0.55 $118,867 350,666 0.61 36 28-43 Up 50 100% 30% $58,049 78 $744
Circle Yukon 500 $0.45 $85,445 347,571 0.54 36 34-35 Even 50 100% 30% $47,057 115 $409
Crooked Creek Kuskokwim 500 $0.51 $114,630 306,349 0.74 31 29-34 Up 50 100% 30% $46,693 131 $356
Fort Yukon Yukon 3,000 $0.35 $125,768 3,384,401 0.11 348 318-380 Up 400 100% 30% $352,032 608 $579
Galena* Yukon 3.60 13.00 1,000 $0.39 $1,368,500 3,918,056 0.90 403 669-100 Down 500 100% 30% $458,442 486 $943
Grayling* Yukon 3.60 9.00 500 $0.50 $143,750 637,308 0.45 65 60-70 Up 100 100% 30% $95,376 142 $672
Holy Cross* Yukon 3.10 10.00 3,500 $0.52 $170,430 516,535 0.64 53 66-60 Down 100 100% 30% $80,372 169 $476
Hughes** Koyukuk 500 $0.28 $75,446 432,056 0.63 44 40-54 Up 75 100% 30% $35,654 71 $503
Huslia* Koyukuk 500 $0.39 $255,300 1,007,445 0.65 104 95-108 Up 150 100% 30% $117,386 224 $524
Igiugig Kvichak 1,000 $0.70 $28,605 252,108 0.16 26 24-27 Up 50 100% 30% $53,242 63 $850
Kalskag* Kuskokwim 3.70 8.00 750 $0.15 $308,200 1,223,304 1.64 126 123-126 Even 150 100% 30% $56,461 255 $221
Kaltag* Yukon 3.50 11.00 500 $0.31 $540,500 799,384 2.16 82 76-87 Up 100 100% 30% $75,009 167 $449
Kiana* Kobuk 1.00 3.92 500 $0.46 $437,000 1,710,873 0.56 176 171-183 Up 200 100% 30% $234,637 391 $600
Koyukuk* Yukon 3.00 9.00 750 $0.30 $12,650 255,000 0.17 27 25-27 Even 50 100% 31% $23,345 80 $292
Manley Hot Springs Tanana 2 mi $0.48 $82,643 272,515 0.64 28 29-27 Down 50 100% 30% $38,948 68 $571
Marshall* Yukon 3.80 8.00 500 $0.37 $331,200 1,436,418 0.62 148 130-165 Up 200 100% 30% $160,476 409 $392
McGrath Kuskokwim 2.80 4.47 500 $0.45 $206,391 2,467,219 0.19 253 285-225 Down 300 100% 30% $333,912 237 $1,409
Mountain Village* Yukon 2.10 7.00 500 $0.31 $709,550 2,774,392 0.83 285 281-291 Up 300 100% 30% $256,870 823 $312
Nikolai Kuskokwim 500 $0.26 $54,787 454,863 0.47 47 45-48 Up 75 100% 30% $35,242 82 $430
Noatak Noatak 0.90 4.49 4,000 $0.67 $508,300 2,109,448 0.36 217 200-247 Up 300 100% 30% $422,645 519 $814
Northway Nabesna 1,000 $0.18 $59,755 1,195,788 0.27 123 143-105 Down 150 100% 30% $66,265 141 $469
Nulato* Yukon 2.80 10.00 500 $0.43 $273,700 997,004 0.64 102 107-95 Down 150 100% 30% $128,925 230 $561
Pilot Station* Yukon 2.50 7.00 1,000 $0.41 $46,000 1,773,393 0.06 182 181-187 Up 200 100% 30% $216,596 612 $354
Red Devil Kuskokwim 500 $0.62 $129,170 145,085 1.43 15 15-15 Even 25 100% 30% $27,177 23 $1,182
Ruby* Yukon 2,000 $0.15 $73,600 669,601 0.72 69 68-68 Even 100 100% 30% $30,758 161 $191
Russian Mission* Yukon 750 $0.38 $224,250 924,302 0.65 95 92-101 Up 150 100% 30% $104,294 354 $295
Shungnak* Kobuk 500 $0.60 $394,450 1,501,060 0.44 154 151-154 Up 200 100% 30% $268,663 272 $988
Slana** Copper 750 $0.32 $97,430 931,425 0.33 96 53-135 Up 150 100% 30% $44,439 94 $473
Sleetmute Kuskokwim 2.90 4.00 500 $0.63 $118,116 329,862 0.57 34 26-40 Up 50 100% 30% $62,131 67 $927
St. Mary's/Pitka's* Yukon 2.70 9.00 3,500 $0.44 $717,600 3,239,575 0.51 333 319-347 Up 400 100% 30% $423,426 842 $503
Stevens Village*/** Yukon 4.00 6.04 1,000 $0.50 $28,825 280,000 0.21 29 29-29 Even 50 100% 30% $41,796 71 $589
Stony River Kuskokwim 2.90 3.24 500 $0.71 $134,565 133,293 1.43 14 13-13 Down 25 100% 30% $28,238 36 $784
Tanana Tanana 1.60 500 $0.43 $592,766 1,233,407 1.13 127 140-121 Down 150 100% 30% $157,404 238 $662
Tok/Dot Lake/Tanacross Tanana 500 $0.33 $1,187,229 11,852,560 0.30 1,218 1245-1204 Down 400 33% 10% $385,638 1742 $221
* Non-Fuel Costs based on highest value of 2008 and 2009 data, plus 15%
**Community PCE data too wide-ranging for realistic projection. These figures are estimates.
NOTES
1. Summer Load Range is the expected change (using PCE data trend to 2020) for monthly load between May 15 to September 15.
2. Percent Summer Percent Offset: Calculated using average of each year's (actual and projected) summer load minus turbine supplement.
3. Per Capita Savings: Calculated using an average of each year's estimated fuel offset, fuel cost, and population.
Summary Table Rivers Village 5 Tables.xlsx Page 1 of 1 Report Date: February 2011
Nationwide River Resource
Assessment
Poncelet Kinetics RHK100 Hydrokinetic Device
Prepared by:
CE2 Engineers, Inc.
Report Date:
March 2011
Communities located throughout the Lower 48 states and Hawaii may potentially benefit by integrating
an RHK100 hydrokinetic device into their existing power system. Potential benefit will depend on the
community’s proximity to water, the estimated stream velocity, the estimated amount of power derived
from the stream, and the value of electricity offset. In each of these states there are several hundred
streams and rivers that may prove adequate for hydrokinetic power generation using the RHK100.
Stream velocity data collected by the United States Geological Survey demonstrate this potential by
state providing an approximation of the value of power offset. Although there appears to be
considerable river energy in each state, the potential benefit of adding an RHK100 to a community’s
existing system remains unclear and will require a community-by-community and a river-by-river
analysis.
Contents
I. Introduction ............................................................................................................................. 1
II. Method .................................................................................................................................... 2
A. List of River Gauge Site Data by State ............................................................................................... 2
B. Residential Rates by State................................................................................................................. 3
III. Results and Discussion ............................................................................................................. 4
A. River Gauge Site Data by State ......................................................................................................... 4
B. Residential Rates by State................................................................................................................. 5
C. Analysis ............................................................................................................................................. 6
List of Appendices
Appendix A—Stream Gauging Sites
Appendix B—Potential Energy (kW) Production
Appendix C—Energy Production and Residential Cost Savings Projections
Appendix D—Residential Rate Projections
CE2 Engineers, Inc. i Report Date: March 2011

I. Introduction
Whitestone Power and Communications (WPC) is in the process of developing and trademarking an
electrical generation device called “the Poncelet Kinetics RHK100”. This River In-Stream Energy
Conversion (RISEC) device can be used in communities located in proximity to a sufficient water
resource to generate electrical power from the hydrokinetic water flow. The RHK100 is a pontoon-
mounted undershot water wheel with a nominal electrical power output capacity of 100 kW. The
preliminary float footprint is estimated at 34 feet long by 19 feet wide, with a weight of approximately
15,000 pounds. The installation will be moored to the shore and protected with Coast Guard-approved
safety equipment.
WPC contracted CE2 Engineers, Inc. (CE2) to perform a preliminary assessment of river energy in the
Lower 48 states. The aim of the preliminary assessment was to identify the potential to extract energy
from rivers using the RHK100. The assessment includes:
1. a summary count of USGS surface water gauging data sites, per state, where average stream
velocity measurements fall between 3 feet/second and 16 feet/second,
2. a summary estimate of potential kilowatts, by velocity range, by state,
3. a summary table of projected residential cost per kilowatt hour, by state,
4. a summary table of potential energy cost savings, by state.
CE2 Engineers, Inc. 1 Report Date: March 2011

II. Method
The four components of the assessment were performed independently. The methods employed are
described here.
A. List of River Gauge Site Data by State
Data describing surface water flow for rivers and streams in the Lower 48 states and Hawaii exist
in several formats on the United States Geological Survey (USGS) National Water Information
System at http://waterdata.usgs.gov/nwis/measurements. The “Field Measurements” database
provided the most comprehensive data set used in this assessment.
USGS NWIS has stream gauge velocity data available for thousands of locations in all 50 states.
Many recordings are taken at sites along minor and major rivers throughout the country.
Multiple measurements are taken at each site, and any number of sites may be situated along a
particular river or stream. Therefore, a “site” refers only to a gauging station along a river. The
site data are downloadable from the NWIS website on a state-by-state basis. The data obtained
for this assessment contained many tens of thousands of data entries per state.
The site velocity data were analyzed through a three-step process. First, all sites measurements
of less than 3 feet/second were filtered out due to an insufficient quantity of economically-
recoverable energy. Then, an average velocity was calculated for each site. The sites were then
categorized and counted based on average velocity measurement.
Identifying the location of the sites was not attempted in this assessment. Each river gauge site
is identified in the database by a “Site Code” that refers to a separate list of site location
descriptions. These data exist in a text format (e.g. “# USGS 03410045 PINE CREEK ABOVE
MOUTH NEAR ONEIDA, TN”) and do not lend themselves easily to either a database or a
geographic analysis.
The site code description was included in the downloadable data records for each state. The
NWIS database does contain county information for each gauging site, but these data are not
available for download. County information, as well as latitude and longitude information for
each site, are only available as “display” when viewing the data for an individual site. For the
purposes of summarizing large amounts of site data for this assessment, the geographical data
for each particular site were not included.

At this preliminary stage no geographical examination of river velocity data was performed.
River names, latitude/longitude coordinates, and county information are available for further
analysis as needed.
B. Residential Rates by State
The U.S. Department of Energy, Energy Information Administration (EIA) maintains a website
with data on various measurements of energy production, consumption, and cost throughout
the nation. A simple cross-tabulation of the data tables downloaded from the EIA website
(http://www.eia.doe.gov/electricity/) provided an overview of the prices residential customers
are paying for electricity.

III. Results and Discussion

A. River Gauge Site Data by State
The table in Appendix A shows a count of the gauging station sites within rivers and streams in
each of the Lower 48 states and Hawaii. Each site has multiple velocity measurements, and the
average velocity was computed.
Each site enumerated in the assembly of this table should not be confused with a single river or
stream. Any single river may have a number of river gauging sites along its length. While every
site listed in this table will not prove to be acceptable for deployment of the RHK100 device, it is
assumed for the purposes of assessing the magnitude of the potential resource, that each site
represents a potential RHK100 device location.
The USGS NWIS database does not describe a particular method of sampling. The standard, if
any, for the placement of river gauging sites is unknown. The site measurements are treated
simply as individual spots on a river where velocity measures were taken.
The number of sites varies by state for obvious reasons of size, geography, and topography.
Some states covering small geographic areas have fewer rivers and streams, and thus fewer data
sites: Delaware has 23, Rhode Island has 40, and Vermont has 49. Louisiana only has 58 sites
with data, possibly because a large percentage of the water in that state moves slowly. The
states with the most sites are California (774), Idaho (629), and New York (564). Generally, the
more sites per state, the more potential energy for the RHK100 to capture.
Site data revealed that the majority of the gauging sites in most states are on streams with an
average velocity ranging between 3 to 4 feet/second. In California, for example, 378 of the 774
sites are in the 3-4 ft/sec range.
In each state, the number of stream sites where the velocity averaged greater than 6
feet/second represents a small portion of the total sites, including California (43 out of 774 =
5.6%), Kansas (4.5%), Massachusetts (4.3%). In Florida, 14.4% of sites had an average velocity of
6 or greater, which points to a sampling pattern where slower rivers had fewer gauging sites
compared to other states.
The table in Appendix B shows the potential amount of energy generated if a 100kW RHK100
device were installed at every data site in every state. Although this is certainly an unrealistic
expectation, the estimate is useful in this energy assessment in showing how the riverine
resources are distributed BY STATE, as well as the distribution of potential energy between
rivers of different velocities.

The amount of energy potentially derived from a river depends on the river’s velocity. In this
assessment, the lowest range stream velocity of 3.0 to 4.0 feet/second was considered the
baseline, where one RHK100 installed at one of these sites would generate 100kW of power.
As the velocity increases by 1 feet/second, the amount of potential extracted energy increases
by a power of 3. With 3.5 feet/second as the baseline where 100kW are generated, each
incremental increase of 1 foot/second was calculated. So, for every site in the 4-5 ft/sec range,
approximately 200% more energy is generated than at the 3-4 ft/sec range. In the 5-6 ft/sec
range, 400% more energy would be generated. At the 9-10 ft/sec range, the increase would be
2000%. And at the top of this table, the 15-16 ft/sec range, 8600% more energy would be
generated than if the same device were installed at the 3-4 ft/sec velocity site.
Some states have enough high-velocity sites to potentially generate a large amount of
electricity. North Carolina has two (2) sites in the 15-16 ft/sec range, as does Florida, Illinois, and
Texas.
Using this formula to estimate potential energy generation at stream sites in each of the velocity
ranges from 3 to 16 feet/second, an estimate for the total potential energy was created BY
STATE. The states with the lowest kilowatt generation potential are Rhode Island (5,200),
Vermont (8,400), and New Hampshire (17,300). Other low-potential states are New Hampshire
(17,300), Maine (19,700), and Delaware (23,200).
The states with the highest kilowatt generation potential are California (198,000), New York
(176,600), Colorado (116,800), and Pennsylvania (108,900).
B. Residential Rates by State
The Energy Information Administration publishes residential electricity rate data for each of the
United States from 1999 to 2009. These figures were used to compute a “least squares” trend
line, a statistical instrument useful for observing mathematical trends in the real world and
using those to predict future trends. Residential electric costs, per state, were projected through
the year 2030.
Appendix C shows energy production and residential cost saving projections for each state.
Currently, customers in New York, Connecticut, Massachusetts and Rhode Island pay the most
among the contiguous 48 states. Hawaii is highest overall. The 2010 costs (in cents) and
projected costs for 2020 and 2030 are shown in the matrix on the next page.

2010 Rate 2020 Rate 2030 Rate

State (Projected) (Projected) (Projected)
Hawaii 25.8 34.4 43.0
New York 17.3 20.1 22.9
Connecticut 17.3 21.4 25.5
Massachusetts 15.9 19.3 22.7
Rhode Island 14.8 17.3 19.8
The cost of electricity is not expected to rise above 10 cents per kilowatt hour (kWh) in a
number of states, including: Arkansas, Arizona, Idaho, Kansas, Kentucky, Nebraska, New Mexico,
and South Dakota. Costs will remain below 9 cents in Illinois, Missouri, North Dakota, Utah and
West Virginia.
Appendix D shows the projected costs for residential electricity per kWh, per state. These
figures are used to calculate the potential benefit of using an RHK100 to capture hydrokinetic
power, as reflected in cost savings to residential customers.
C. Analysis
The data published by the USGS and the EIA provide a basis for estimating both the potential
river and stream energy in each of the Lower 48 states and Hawaii, and the potential cost
savings of capturing stream energy using an RHK100 hydrokinetic device. To recap, the analysis
represents a situation where an RHK100 is deployed at each and every gauging site reported in
the USGS database, after eliminating sites where velocity measurements were under 3
feet/second. Nothing was done to determine a site’s viability, and it is certain that many sites
included in this assessment may be either difficult or impossible to equip with an RHK100.
Reasons include terrain (steepness, narrowness), volume, depth, seasonal ice, proximity to a
community, environmental impact, shipping, or recreational use, among others. None of these
variables are considered in this assessment.
The cost savings figures, likewise, do not take into consideration the cost of building, deploying,
or operating/maintaining an RHK100. They reflect only the high-end potential deferred cost of
deploying the maximum number of RHK100s in any given state.
The table in Appendix C shows the estimated projected potential cost savings by state.
California, with its middle-high residential energy rates and its large number of potential
RHK100 locations (with a higher percentage of those in fast-moving waters), could potentially
benefit the most from installation of the RHK100 hydrokinetic device. Any one site in California
could save the equivalent of $31.6 million in 2010 residential rate dollars. With 774 RHK100s
deployed throughout the entire state, California could save $24.5 billion (in 2010 residential rate

dollars). With the projected increase in residential rates, these potential savings rise to $40.5
million per site, and $31.4 billion statewide, in 2030 (Appendix D).
The following matrix, data taken from Appendix D, shows the five states with the highest and
lowest potential savings, by site and by state.
State 2010 2020 2030

Highest Savings Per Site (in millions
Delaware $106.3 $125.8 $145.3
Hawaii $55.2 $73.7 $92.1
Florida $45.1 $53.8 $62.5
New York $47.4 $55.1 $62.8
Texas $38.5 $47.7 $57.0
Lowest Savings Per Site (in millions)
Michigan $12.2 $13.8 $15.3
South Dakota $12.5 $13.6 $14.7
Missouri $13.2 $13.7 $14.1
North Dakota $13.2 $14.4 $15.6
Kansas $14.8 $15.5 $16.3
Highest Savings Per State (in billions)
California $24.5 $27.9 $31.4
New York $26.8 $31.1 $35.4
Texas $17.2 $21.3 $25.4
Idaho $10.8 $12.7 $14.6
Colorado $9.7 $11.2 $12.7
Lowest By State (In billions)
Rhode Island $0.7 $0.8 $0.9
Vermont $1.1 $1.3 $1.5
Louisiana $1.6 $1.7 $1.9
North Dakota $2.0 $2.2 $2.4
West Virginia $2.3 $2.4 $2.6
While the analysis of the EIA data showed the cost of residential power rising in all states, some
states will rise faster than others. That difference in rate of increase affects the annual fixed
RHK100 cost savings, and causes some states’ cost savings to increase more rapidly than others.

In sum, rivers and streams in the Lower 48 states and Hawaii may potentially contain vast
amounts of moving water where an RHK100 hydrokinetic device would successfully generate
electricity. The estimates used in this report represent a possible maximum amount of energy
generated by the RHK100 in these rivers and streams (with a velocity over 3 feet/second), and a
broad comparison of potential cost savings between states.
The estimates should only serve as a general guideline to each state’s potential. Data were not
screened or corrected for issues relating to sampling bias, river dimensions, seasonal ice,
proximity to a community power plant, boat traffic conditions, recreation, or environmental
sensitivity, and any particular proposal to develop a hydrokinetic device would have to take all
these into close consideration.
The cost savings estimates do not include any accounting for the cost of engineering, building,
deploying, operating, or maintaining the RHK100 device. Those capital and operational costs
would reduce the cost savings reported here.

Appendix A—Stream Gauging Sites
CE2 Engineers, Inc. Report Date: March 2011

STREAM GAUGING SITES Appendix A
GROUPED BY AVERAGE VELOCITY
Average
Feet/Second AL AR AZ CA CO CT DE FL GA HI IA ID IL IN KS KY LA MA MD ME MI
3-4 128 106 97 378 278 43 5 115 324 37 223 318 146 129 128 175 37 89 75 33 127
4-5 50 64 72 256 187 49 7 44 117 18 125 218 49 51 53 92 17 38 53 40 51
5-6 12 20 34 97 63 7 3 14 36 10 31 55 6 7 9 22 8 18 21 7
6-7 5 8 8 20 10 4 2 7 11 8 15 12 2 4 4 6 2 6
7-8 3 4 3 6 6 4 2 5 9 6 11 4 3 2 11 1 3 5
8-9 2 6 1 1 3 4 4 4 1 1 1 3 1 1
9-10 2 1 3 2 4 1 1 2 2 1 3 1 1
10-11 1 1 1 2 1 1 2 1 1 1 3
11-12 2 3 3 1 2 2 1 1 1
12-13 1 2 3 1 1 3 2 1
13-14 2 2 1 1 2 1 1 1 1 1
14-15 1 1 2 1 1 1 2 3
15-16 1 1 1 2 1 1 2
State Total 204 211 215 774 546 112 23 202 507 74 412 629 216 201 199 313 58 141 164 94 185
Appendix A Page 1 of 3
Average
Feet/Second MN MO MS MT NC ND NE NH NJ NM NV NY OH OK OR PA RI SC SD TN TX
3-4 177 203 111 203 139 94 125 29 147 106 159 240 112 109 137 303 30 87 135 133 239
4-5 89 112 54 112 76 40 47 32 71 72 62 199 60 71 141 132 9 18 65 81 126
5-6 20 46 12 46 29 8 5 14 18 10 16 73 13 15 48 27 1 7 3 24 38
6-7 11 4 4 4 12 4 1 4 11 5 7 22 3 7 10 9 2 6 9 12
7-8 3 2 4 2 4 5 1 4 2 3 9 2 3 3 2 5 6 13
8-9 3 3 1 3 8 1 1 1
9-10 3 2 3 2 1 3 2 1 1 1 5
10-11 3 1 1 1 2 4 1 3 1 1 4
11-12 2 1 2 1 2 1 1 1 1 1
12-13 2 2 1 1 1
13-14 1 1 1 1 4 1 1 2
14-15 1 1 1 1 1 1 3
15-16 2 1 1 1 2
State Total 309 369 192 369 271 152 182 79 258 201 252 564 196 206 337 481 40 119 215 257 446
Average
Feet/Second UT VA VT WA WI WV WY State Total
3-4 147 165 24 144 180 52 150 6871
4-5 88 78 20 157 64 48 98 3873
5-6 33 27 5 64 18 28 15 1143
6-7 21 6 15 6 4 5 338
7-8 1 1 4 4 2 173
8-9 3 1 3 1 62
9-10 1 1 50
10-11 1 1 39
11-12 29
12-13 1 1 23
13-14 25
14-15 1 1 23
15-16 16
State Total 294 278 49 388 276 136 269 12665
Appendix B—Potential Energy (kW) Production

POTENTIAL ENERGY (kW) PRODUCTION Appendix B
GROUPED BY VELOCITY
(BASED ON 100kW PER 3.5 FEET/SECOND)
Average
Feet/Second AL AR AZ CA CO CT DE
3-4 12,800 10,600 9,700 37,800 27,800 4,300 500
4-5 10,000 12,800 14,400 51,200 37,400 9,800 1,400
5-6 4,800 8,000 13,600 38,800 25,200 2,800 1,200
6-7 3,000 4,800 4,800 12,000 6,000 2,400 1,200
7-8 3,000 4,000 3,000 6,000 6,000 4,000 2,000
8-9 2,800 - - 8,400 - 1,400 1,400
9-10 - 4,000 2,000 6,000 - 4,000 -
10-11 - 2,700 - 2,700 - - 2,700
11-12 - 7,000 - 10,500 - - -
12-13 - 4,500 - 9,000 - - -
13-14 11,400 11,400 - - - 5,700 5,700
14-15 7,100 - - 7,100 14,200 7,100 7,100
15-16 8,600 8,600 - 8,600 - - -
Total Potential kW 63,500 78,400 47,500 198,100 116,600 41,500 23,200
Total Number of State Sites 204 211 215 774 546 112 23
kW Per Site (B33/B34) 311 372 221 256 214 371 1,009
Total kWh Year Site (B52*8760) 2,726,765 3,254,900 1,935,349 2,242,062 1,870,725 3,245,893 8,836,174
Total kWh Year State (B33*8760) 556,260,000 686,784,000 416,100,000 1,735,356,000 1,021,416,000 363,540,000 203,232,000
Appendix B Page 1 of 8
GROUPED BY VELOCITY
Average
Feet/Second FL GA HI IA ID IL IN
3-4 11,500 32,400 3,700 22,300 31,800 14,600 12,900
4-5 8,800 23,400 3,600 25,000 43,600 9,800 10,200
5-6 5,600 14,400 4,000 12,400 22,000 2,400 2,800
6-7 4,200 6,600 4,800 9,000 7,200 1,200 2,400
7-8 5,000 9,000 - 6,000 11,000 4,000 3,000
8-9 4,200 5,600 - 5,600 5,600 1,400 1,400
9-10 8,000 2,000 2,000 4,000 4,000 2,000 6,000
10-11 5,400 2,700 - 2,700 - 5,400 -
11-12 10,500 3,500 - 7,000 7,000 - -
12-13 13,500 4,500 - 4,500 13,500 9,000 -
13-14 - 11,400 - - 5,700 5,700 -
14-15 - - - 7,100 14,200 - 21,300
15-16 17,200 - - 8,600 8,600 17,200 -
kW Per Site (B33/B34) 465 228 245 277 277 337 299
Total kWh Year State (B33*8760) 822,564,000 1,011,780,000 158,556,000 1,000,392,000 1,525,992,000 636,852,000 525,600,000
GROUPED BY VELOCITY
Average
Feet/Second KS KY LA MA MD ME MI
3-4 12,800 17,500 3,700 8,900 7,500 3,300 12,700
4-5 10,600 18,400 3,400 7,600 10,600 8,000 10,200
5-6 3,600 8,800 - 3,200 7,200 8,400 2,800
6-7 2,400 3,600 - 1,200 3,600 - -
7-8 2,000 11,000 1,000 3,000 5,000 - -
8-9 1,400 4,200 - 1,400 1,400 - -
9-10 - 2,000 - - 2,000 - -
10-11 2,700 2,700 2,700 - 8,100 - -
11-12 - 3,500 3,500 - 3,500 - -
12-13 4,500 - - - - - -
13-14 - 5,700 5,700 - 5,700 - -
14-15 - - - - - - -
15-16 - - - - - - -
kW Per Site (B33/B34) 201 247 345 179 333 210 139
Total kWh Year State (B33*8760) 350,400,000 678,024,000 175,200,000 221,628,000 478,296,000 172,572,000 225,132,000
GROUPED BY VELOCITY
Average
Feet/Second MN MO MS MT NC ND NE
3-4 17,700 20,300 11,100 20,300 13,900 9,400 12,500
4-5 17,800 22,400 10,800 22,400 15,200 8,000 9,400
5-6 8,000 18,400 4,800 18,400 11,600 3,200 2,000
6-7 6,600 2,400 2,400 2,400 7,200 2,400 600
7-8 3,000 2,000 4,000 2,000 4,000 5,000 1,000
8-9 4,200 - - - 4,200 - 1,400
9-10 - - 6,000 - 4,000 - -
10-11 8,100 2,700 - 2,700 - - 2,700
11-12 7,000 - 3,500 - 7,000 3,500 -
12-13 - - 9,000 - 9,000 - -
13-14 5,700 5,700 - 5,700 - - -
14-15 - - 7,100 - - - 7,100
15-16 - - - - 17,200 - -
kW Per Site (B33/B34) 253 200 306 200 344 207 202
GROUPED BY VELOCITY
Average
Feet/Second NH NJ NM NV NY OH OK
3-4 2,900 14,700 10,600 15,900 24,000 11,200 10,900
4-5 6,400 14,200 14,400 12,400 39,800 12,000 14,200
5-6 5,600 7,200 4,000 6,400 29,200 5,200 6,000
6-7 2,400 6,600 3,000 4,200 13,200 1,800 4,200
7-8 - 4,000 2,000 3,000 9,000 2,000 3,000
8-9 - 4,200 - - 11,200 1,400 -
9-10 - 6,000 4,000 2,000 6,000 4,000 -
10-11 - - - 5,400 10,800 2,700 -
11-12 - - 7,000 3,500 3,500 3,500 -
12-13 - - 4,500 - - - -
13-14 - 5,700 - - 22,800 - 5,700
14-15 - - - 7,100 7,100 - -
15-16 - - 8,600 - - 8,600 -
kW Per Site (B33/B34) 219 243 289 238 313 267 214
Total kWh Year State (B33*8760) 151,548,000 548,376,000 508,956,000 524,724,000 1,547,016,000 459,024,000 385,440,000
GROUPED BY VELOCITY
Average
Feet/Second OR PA RI SC SD TN TX
3-4 13,700 30,300 3,000 8,700 13,500 13,300 23,900
4-5 28,200 26,400 1,800 3,600 13,000 16,200 25,200
5-6 19,200 10,800 400 2,800 1,200 9,600 15,200
6-7 6,000 5,400 - 1,200 3,600 5,400 7,200
7-8 - 3,000 - 2,000 5,000 6,000 13,000
8-9 - - - - 1,400 - 1,400
9-10 2,000 - - 2,000 - 2,000 10,000
10-11 - 8,100 - 2,700 - 2,700 10,800
11-12 - 3,500 - - - 3,500 -
12-13 - - - - - 4,500 4,500
13-14 - 5,700 - - - - 11,400
14-15 - 7,100 - 7,100 - - 21,300
15-16 - 8,600 - - - - 17,200
kW Per Site (B33/B34) 205 226 130 253 175 246 361
Total kWh Year State (B33*8760) 605,316,000 953,964,000 45,552,000 263,676,000 330,252,000 553,632,000 1,411,236,000
GROUPED BY VELOCITY
Average
Feet/Second UT VA VT WA WI WV WY
3-4 14,700 16,500 2,400 14,400 18,000 5,200 15,000
4-5 17,600 15,600 4,000 31,400 12,800 9,600 19,600
5-6 13,200 10,800 2,000 25,600 7,200 11,200 6,000
6-7 12,600 3,600 - 9,000 3,600 2,400 3,000
7-8 1,000 1,000 - 4,000 4,000 2,000 -
8-9 4,200 1,400 - 4,200 1,400 - -
9-10 - - - 2,000 2,000 - -
10-11 - - - - - 2,700 2,700
11-12 - - - - - - -
12-13 - - - - 4,500 4,500 -
13-14 - - - - - - -
14-15 7,100 - - - 7,100 - -
15-16 - - - - - - -
kW Per Site (B33/B34) 239 176 171 234 220 276 172
GROUPED BY VELOCITY
Average
Feet/Second All States
3-4 687,100
4-5 774,600
5-6 457,200
6-7 202,800
7-8 173,000
8-9 86,800
9-10 100,000
10-11 105,300
11-12 101,500
12-13 103,500
13-14 142,500
14-15 163,300
15-16 137,600
Total Potential kW 3,235,200
Total Number of State Sites 12,665
kW Per Site (B33/B34) 255
Total kWh Year Site (B52*8760) 2,237,691
Total kWh Year State (B33*8760) 28,340,352,000
Appendix C—Energy Production and Residential Cost Savings Projections

ENERGY PRODUCTION AND RESIDENTIAL COST SAVINGS PROJECTIONS Appendix C
(BASED ON PRODUCTION RATE OF 100kW PER 3.5 FEET/SECOND)
Description AL AR AZ CA CO CT DE
kW Per Site (B33/B34) 311 372 221 256 214 371 1,009
Total kWh Year State (B33*8760) 556,260,000 686,784,000 416,100,000 1,735,356,000 1,021,416,000 363,540,000 203,232,000
Est. Price per kWh (cents) 2010 9.40 8.35 9.32 14.10 9.50 17.28 12.04
Potential Savings Per Site, 2010 $25,625,274 $27,174,650 $18,032,460 $31,613,546 $17,765,687 $56,091,591 $106,364,281
Potential Savings Per State, 2010 $5.2 billion $5.7 billion $3.9 billion $24.5 billion $9.7 billion $6.3 billion $2.4 billion
Appendix C Page 1 of 8
Description FL GA HI IA ID IL IN
kW Per Site (B33/B34) 465 228 245 277 277 337 299
Total kWh Year State (B33*8760) 822,564,000 1,011,780,000 158,556,000 1,000,392,000 1,525,992,000 636,852,000 525,600,000
Description KS KY LA MA MD ME MI
kW Per Site (B33/B34) 201 247 345 179 333 210 139
Description MN MO MS MT NC ND NE
kW Per Site (B33/B34) 253 200 306 200 344 207 202
Description NH NJ NM NV NY OH OK
kW Per Site (B33/B34) 219 243 289 238 313 267 214
Total kWh Year State (B33*8760) 151,548,000 548,376,000 508,956,000 524,724,000 1,547,016,000 459,024,000 385,440,000
Description OR PA RI SC SD TN TX
kW Per Site (B33/B34) 205 226 130 253 175 246 361
Total kWh Year State (B33*8760) 605,316,000 953,964,000 45,552,000 263,676,000 330,252,000 553,632,000 1,411,236,000
Description UT VA VT WA WI WV WY
kW Per Site (B33/B34) 239 176 171 234 220 276 172
Description All States

Total Potential kW 3,235,200
Total Number of State Sites 12,665
kW Per Site (B33/B34) 255
Total kWh Year Site (B52*8760) 2,237,691
Total kWh Year State (B33*8760) 28,340,352,000
Est. Price per kWh (cents) 2010
Potential Savings Per Site, 2010 $1,232,498,247
Potential Savings Per State, 2010 $294.0 billion
Appendix D—Residential Rate Projections

Residential Rate Projections Appendix D
based on EIA data from 1990-2009
Rates are expressed in cents per kWh
Residential Rates Year

State 1990 2000 2010 2020 2030
AL 6.59 7.05 9.40 11.16 12.93
AR 8.07 7.45 8.35 8.67 8.99
AZ 9.04 8.44 9.32 9.55 9.78
CA 9.98 10.89 14.10 16.10 18.09
CO 7.02 7.31 9.50 10.96 12.42
CT 10.01 10.86 17.28 21.40 25.52
DE 8.39 8.54 12.04 14.24 16.45
FL 7.77 7.77 11.09 13.23 15.36
GA 7.46 7.60 9.18 10.21 11.24
HI 10.26 16.41 25.79 34.40 43.01
IA 7.81 8.37 9.65 10.68 11.70
ID 4.87 5.39 7.06 8.30 9.54
IL 9.92 8.83 9.26 8.92 8.59
IN 6.87 6.87 8.41 9.47 10.54
KS 7.83 7.65 8.41 8.84 9.27
KY 5.69 5.47 7.27 8.37 9.47
LA 7.41 7.67 8.97 9.97 10.96
MA 9.66 10.53 15.86 19.27 22.68
MD 7.22 7.95 11.43 13.78 16.13
ME 9.30 12.49 15.47 17.96 20.46
MI 7.83 8.52 10.11 11.36 12.62
MN 6.80 7.52 9.25 10.67 12.08
MO 7.36 7.04 7.57 7.81 8.06
MS 6.89 6.93 9.57 11.27 12.98
MT 5.45 6.49 9.00 10.91 12.82
NC 7.84 7.97 9.28 10.11 10.94
ND 6.26 6.44 7.30 7.95 8.61
NE 6.23 6.53 7.78 8.75 9.72
NH 10.34 13.15 15.18 17.13 19.08
NJ 10.36 10.27 13.73 15.50 17.28
NM 8.94 8.36 9.23 9.44 9.66
NV 5.70 7.28 12.18 15.74 19.30
NY 11.44 13.97 17.31 20.10 22.90
OH 8.05 8.61 9.58 10.39 11.20
OK 6.58 7.03 8.44 9.48 10.51
OR 4.73 5.88 8.50 10.56 12.62
PA 9.22 9.53 10.74 11.54 12.33
RI 9.84 11.28 14.80 17.29 19.78
SC 7.15 7.58 9.46 10.84 12.22
Appendix D Page 1 of 2
Residential Rate Projections Appendix D
based on EIA data from 1990-2009
Rates are expressed in cents per kWh
Residential Rates Year

State 1990 2000 2010 2020 2030
SD 6.95 7.42 8.19 8.91 9.62
TN 5.69 6.33 8.28 9.86 11.44
TX 7.20 7.96 12.18 15.10 18.02
UT 7.13 6.29 7.77 8.38 8.98
VA 7.25 7.52 9.11 10.16 11.20
VT 9.27 12.30 14.94 17.83 20.71
WA 4.39 5.13 7.53 9.30 11.06
WI 6.63 7.53 11.08 13.75 16.42
WV 5.90 6.27 6.87 7.34 7.80
WY 5.97 6.50 8.14 9.44 10.74
Maximum Rate 11.44 16.41 25.79 34.40 43.01
Note: 1990 and 2000 data are unprojected costs reported by EIA.
The 2010, 2020, and 2030 rates are projected.
Appendix D Page 2 of 2

Final Project Report

Uploaded by

Copyright:

Available Formats

Final Project Report

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Final Project Report

Uploaded by

Copyright:

Available Formats

DE-EE0004573

Whitestone Poncelet RISEC Project

Final Project Report

Project Title: Whitestone Poncelet RISEC Project

Recipient: Whitestone Power and Communications

DOE Project Team: DOE HQ Program Manager – Jacques Beaudry-Losique

Executive Summary ......................................................................................................................3

Effectiveness and Feasibility

1. Efficiency Paradigm - This project analyzed attempted applications of RISEC

Overall Design Criteria

Figure 1: Early embodiment of RISEC device

Figure 2: Early Embodiment of RISEC device

A number of factors contributed to a significant paradigm shift regarding power generation. It

Figure 3: Final embodiment of RISEC device

Power Generation/ Conditioning

Electrics/ Controls/ Monitoring

This may be related to the angle of heel by the following equation:

R.M .    Volumedisplaced  Height metacentric  

Weight1  Height1  Weight 2  Height 2  ...  Weight n  Height n

A simplified center of gravity for craft + wheel was estimated as follows:

Thus a simplified estimate of CG is 4.75 feet.

Metacentric Height is calculated as follows:

194.47 - 4.75 + 2.68 = 192.4 feet

and bending stresses would be expressed:

Material Properties Table

Figure7: ISO clipping- stress <2000 psi

Figure 8: ISO clipping- stress > 102 psi

The Strouhal/Reynolds relation is graphically presented in the following figure:

Figure 9: Chart relating dimensionless parameters for a cylinder in cross-flow

The vortex shedding frequency is described by the following equation:

Load Strengthcompressive _ yield

ec is the eccentricity (for this calculation it was assumed to be five inches)

E is the elastic modulus.

Figure 10: FEA model of rigid struts

Figure11: Results of FEA model

Blade Design Calculations

The power developed by an undershot waterwheel in unconfined flow is expressed4:

Blade Design Process

Figure 12: Early blade design with aluminum ribs

Figure 13: Detail of paddle with detente notch mount

Figure 14: Early all-plastic blade design

Consequently, considerable research was conducted to determine a suitable material which

Material Elastic Modulus Yield Strength % Elongation at Machinability

Force  Velocitylog   masslog k paddle

b1h13 b2 h23 b3 h33

The area moment of inertia was calculated to be:

The stiffness was then expressed:

These values may be substituted to determine the maximum force:

To determine maximum stress in paddle, the following equation is employed:

Substituting these values:

Figure 15: Comparison closed form model to actual model

A static test was conducted with the following parameters:

Figure 16: FEA results of paddle loading

Figure 17: Testing jig with springs

Figure 18: Log strike test- note cracking at weld seam.

Figure 19: Blade deformation recovery after log strike test

Figure 20: Water flow test setup

Figure 21: Blade under water flow test

Figure22: Deformed spring under water flow test