Attachment Holland:DAW 1-112
Exh. Pet.-ML-3
ICE THROW RISK ASSESSMENT FOR THE PROPOSED KINGDOM COMMUNITY WIND POWER PROJECT
Client Green Mountain Power Corp. Contact Charles Pughe Document No 41404/AR/02A Classification Client’s Discretion Status Final Date 17 November 2010
Authors: Y Boucetta / P Heraud
Checked by: MLeblanc
Approved by: B Ait-Driss
Holland-DAW_1-112-000 Page 1 Attachment Holland:DAW 1-112
Garrad Hassan America, Inc. Document : 41404/AR/02A
IMPORTANT NOTICE AND DISCLAIMER
This report is intended for the use of the Client on whose instructions it has been prepared, and who has entered into a written agreement directly with Garrad Hassan America, Inc. (“GH”). GH’s liability to the Client is set out in that agreement. GH shall have no liability to third parties for any use whatsoever without the express written authority of GH. The report may only be reproduced and circulated in accordance with the Document Classification and associated conditions stipulated in this report, and may not be disclosed in any public offering memorandum without the express written consent of GH.
This report has been produced from information relating to dates and periods referred to in this report. The report does not imply that any information is not subject to change.
Key To Document Classification
Strictly Confidential : Recipients only
Private and Confidential : For disclosure to individuals directly concerned within the recipient’s organisation
Commercial in Confidence : Not to be disclosed outside the recipient’s organisation
GH only : Not to be disclosed to non GH staff
Client’s Discretion : Distribution at the discretion of the client subject to contractual agreement
Published : Available to the general public
© 2009 Garrad Hassan America, Inc.
Holland-DAW_1-112-000 Page 2 Attachment Holland:DAW 1-112
Garrad Hassan America, Inc. Document : 41404/AR/02A
CONTENTS
Page
1 INTRODUCTION 1
2 ASSESSMENT SUBJECT 2
3 ICE THROW ASSESSMENT METHODOLOGY 4
4 DATA SOURCES AND OTHER INPUTS 6 4.1 Wind climate during icing events 6 4.2 Control methodologies 6 4.3 Assessment guidelines and data 7
5 RESULTS OF ICE THROW ASSESSMENT 9 5.1 Wind turbine icing 9 5.2 Technical feasibility of icing during operation 9 5.3 Individual risk 10 5.4 Control mitigation 16
6 CONCLUSIONS 17
REFERENCES 18
LIST OF TABLES 19
LIST OF FIGURES 20
Holland-DAW_1-112-000 Page 3 Attachment Holland:DAW 1-112
Garrad Hassan America, Inc. Document : 41404/AR/02A
1 INTRODUCTION
Garrad Hassan America, Inc. (“GH”) has been contracted by Green Mountain Power Corporation (the “Client”) to undertake an assessment of the risk of ice fragments shed from wind turbines striking members of the public in the vicinity of some turbines from the proposed Kingdom Community Wind power project (the “Project”).
The results of this work are presented in this report.
1
Holland-DAW_1-112-000 Page 4 Attachment Holland:DAW 1-112
Garrad Hassan America, Inc. Document : 41404/AR/02A
2 ASSESSMENT SUBJECT
The proposed Project site is located Southwest of Orleans, Vermont. The site elevation is approximately between 680 m to 800 m. The Project consists of 21 wind turbines. As there is still uncertainty regarding the wind turbine models of the Project, The Ice Throw Analysis has been performed for three models of wind turbine, namely Siemens SWT-101, Vestas V112 and GE 2.75-100 model. The worst case scenario with the Vestas V112, the turbine model with the largest rotor diameter, is presented in this report.
The key parameters of the wind turbine model are summarized in Table 2.1.
Wind turbine model Vestas V112 RatedPower 3MW Rotor diameter 112 m Hubheight 84m Cut-in wind speed 3 m/s Cut-out wind speed 25 m/s Nominal rotor speed 17.7 rpm Nominal tip speed 103.7 m/s
Table 2.1 Wind turbine parameters
This assessment is focused primarily on the area surrounding each turbine as presented in Figure 2.1. Of particular interest is the risk of ice throw on the surrounding parcels from wind turbines 1, 5, 6, 8, 15, 19, 20 and 21, which are the closest to non-participating landowner property lines.
2
Holland-DAW_1-112-000 Page 5 Attachment Holland:DAW 1-112
Garrad Hassan America, Inc. Document : 41404/AR/02A
Figure 2.1 Proposed locations and property line
3
Holland-DAW_1-112-000 Page 6 Attachment Holland:DAW 1-112
Garrad Hassan America, Inc. Document : 41404/AR/02A
3 ICE THROW ASSESSMENT METHODOLOGY
The assessment methodology used herein is based on that developed by GH in conjunction with the Finnish Meteorological Institute and Deutsches Windenergie-Institut as part of a research project on the implementation of wind energy in cold climates (WECO) primarily funded by the European Union and also supported in part by the United Kingdom Department of Trade and Industry [1]. The guidelines for safety assessments in relation to ice throw were developed by GH in the WECO project and that work was summarized in a series of conference papers [2, 3, and 4]. These guidelines have been applied to the Project site by considering the proposed turbine type, the terrain of the site and surrounding area, and assumptions for human presence in the surrounding area.
The overall approach is presented schematically in Figure 3.1 and is based on a staged approach:
Determine the periods when ice accretion on structures is technically possible, based on historical climatic observations.
Within those periods, determine when the wind speed conditions are within the operational range of the wind turbines.
Within the resultant periods, if applicable, exclude those periods when the wind turbines will be shut down automatically by the wind turbine control system or by remote operators.
Based on an estimate from the above of the amount of icing, use guidelines to arrive at probability of fragments landing at the distances from the turbines which are of interest.
Where information is available, estimate probability of members of the public being present within the distances from the turbine which are being considered.
Arrive at combined probability of members of the public being hit by ice fragments.
Compare that probability to a suitable benchmark risk such as natural hazards.
It is our professional opinion that this methodology is sound and provides an appropriate analysis of the Project.
4
Holland-DAW_1-112-000 Page 7 Attachment Holland:DAW 1-112
Garrad Hassan America, Inc. Document : 41404/AR/02A
Historical reference station data
Temperature Hours when icing Humidity conditions are Adjust to site Cloud cover present Cloud base Precipitation
Observed evidence Validation from area (possibly anecdotal)
Subset of hours when wind speed is Concurrent wind Adjust to site in turbine operating speed range
Likelihood of unplanned shutdown due to ice (mainly sensor icing) Subset of hours Specification of when turbine is turbine and "ice" operating control system Shutdown by ice prevention / detection control system Assess risk of ice throw at distance of interest
Estimate probability of public presence
Estimate risk of Revise control people being struck strategy by ice fragments
Evidence of fragment Acceptable risk? size and mass No Yes
No further action
Figure 3.1 Ice throw risk assessment procedure
5
Holland-DAW_1-112-000 Page 8 Attachment Holland:DAW 1-112
Garrad Hassan America, Inc. Document : 41404/AR/02A
4 DATA SOURCES AND OTHER INPUTS
4.1 Wind climate during icing events
Climatic data recorded at the site during icing events have been provided to GH by the client. The data was recorded for each 10-minute period from sensors mounted on the meteorological tower (mast 808) from June 2003 through May 2008. A joint wind speed and wind direction table of the icing period (November to March) was derived from these measurements and used as the base meteorological input for this study.
4.2 Control methodologies
Ice detectors are typically mounted to the nacelle of a turbine or nearby meteorological tower and monitored by the wind farm control system, triggering an automatic or remote manual shutdown of the wind farm in the event that icing conditions are detected.
It is also generally accepted in the wind industry that any ice build up on the blades of an operating turbine will lead to additional vibration. This is caused by both mass and aerodynamic imbalances. All machines, including the Vestas V112, are equipped with vibration monitors, which will shut the machine down during these periods.
Depending on the ice throw risk, it may be appropriate to implement a winter operating protocol that will curtail the operating of wind turbines in the event of icing and when extreme weather conditions present unsafe conditions for the general public. This typically involves operator or automatic system shut down under one or more of the following circumstances:
The installed ice monitoring device(s) and heated wind sensors (installation subject to reliability testing) detect unsafe conditions are present due to icing conditions;
Ice accretion is recognized by the remote or on site operator;
Air temperature, relative humidity and other meteorological conditions at the site are conducive to ice formation;
Air temperature is several degrees above 0 degrees C after icing conditions; and
Any other weather conditions which appear unsafe.
During any of these events, a typical operating protocol will provide that turbines which present a safety risk to the public are to be placed in Pause mode, in which the units are inoperative.
6
Holland-DAW_1-112-000 Page 9 Attachment Holland:DAW 1-112
Garrad Hassan America, Inc. Document : 41404/AR/02A
4.3 Assessment guidelines and data
The guidelines produced in the WECO project were based on a combination of numerical modeling and observations.
The numerical modeling involved Monte-Carlo simulations of a range of scenarios of ice building up on a wind turbine and being shed from the rotor blades. An updated set of simulations have been conducted for the Project study using the wind turbine parameters of the Vestas V112 model as defined in Table 2.1 and wind regime measured at the site for the period from November to March.
In the modeling, further assumptions are required in regard to the aerodynamic properties of ice fragments. These assumptions were verified during the course of the WECO project by measuring the lift and drag characteristics of models of typical ice fragments in wind tunnels. Those coherent fragments collected from various icing events were irregular blocks shed from the leading edge of the rotor blades. Moulds were produced from these and replicas cast for wind tunnel testing. No stable lifting situation was measured, leading to a conclusion that the lift coefficient could be ignored. The drag coefficient meanwhile was measured to fall in the same range as was assumed in the modeling described above.
In the EU WECO study, the observations of ice build-up on rotor blades and fragments shed from rotor blades were gathered from wind farms throughout Europe. The data gathered are presented in Figure 4.1, which shows that fragments typically land within 100 m of the wind turbine. Ice fragments with masses up to 1 kg (2.2 lbs) were found, although most were much smaller.
100 90 80 70 60 50 40 Ice throw [m] 30 20 10 0 0 102030405060
Turbine Rotor Diameter [m]
Figure 4.1 Recorded ice throw data (from [4])
7
Holland-DAW_1-112-000 Page 10 Attachment Holland:DAW 1-112
Garrad Hassan America, Inc. Document : 41404/AR/02A
The results from the Monte Carlo analysis are shown in Figures 4.2 and 4.3 for 1 kg ice fragments for each 30 degree direction sector. These represent the probabilities, given an ice fragment has been released, that any one ice fragment lands in one square meter of ground area, as a function of distance and direction from the turbine. The results shown in Figures 4.2 and 4.3 were used in risk assessment at the Project site where detailed assessment is required.
0.01
r 0.01 Note: Each line represents a 30° direction sector 0.001 0.001
0.0001 0.0001
0.00001 0.00001
0.000001 0.000001
0.0000001 Probability per square meter per yea 0.0000001
0.00000001 0.000000010 50 100 150 200 250 300 0 50 100 150 200 250 Distance from turbine [m] Distance from turbine [m] Figure 4.2 Calculated probabilities of 1 kg ice fragment throw distances
0.01
0.01 0.001 Note:Each line represents a a 30° direction sector 0.001 0.0001
0.0001 0.00001
0.00001 0.000001
0.000001 0.0000001 Probability per square meter per ye 0.0000001 0.00000001
0.00000001 0 50 100 150 200 250 0 50 100Distance 150 from turbine 200 [m] 250 300 Distance from turbine [m]
Figure 4.3 Calculated probabilities of 1 kg ice fragment drop distances
8
Holland-DAW_1-112-000 Page 11 Attachment Holland:DAW 1-112
Garrad Hassan America, Inc. Document : 41404/AR/02A
5 RESULTS OF ICE THROW ASSESSMENT
5.1 Wind turbine icing
Ice can build up on wind turbine rotor blades when appropriate conditions of temperature and humidity exist, as it will on any structure that is exposed to the elements when appropriate conditions of temperature and humidity exist. When a wind turbine is stationary, it is no more likely to suffer from ice accretion than a large stationary structure such as a building, tree or power line. Similar to such structures, accreted ice will eventually be released and fall directly to the ground.
When operating, which for the Vestas V112 is when the hub height wind speed is in the range 3 m/s to 25 m/s, ice can accrete on the rotor blades in appropriate conditions of temperature and humidity. In this case, observations suggest that higher ice accretion rates occur due to the relative velocity of the rotor blades. Any fragments will land either directly below the wind turbine, in the plane of the wind turbine rotor, or downwind.
When a risk is perceived due to icing of rotor blades, it is common for mitigation measures to be taken in terms of automated or remote manual shutdown of the wind turbines. It is noted that remote monitoring and operation of wind farms is now standard in the industry.
5.2 Technical feasibility of icing during operation
The joint wind speed and direction frequency distribution of the site measurements provided by the Client have been used as the primary source of meteorological data to estimate the technical feasibility of icing during operation.
The Vestas V112 turbines proposed for the Project site do not operate when the hub height wind speeds are outside the 3 m/s to 25 m/s range. Given that ice throw is only feasible when turbines are in operation; conditions outside this wind speed range are not considered.
Based on the above, there is evidence to suggest that 23 days per year may lead to icing conditions. This range and GH previous assessment experience in Vermont indicate that 25 days of icing per year is a representative assumption of the Project site conditions for icing events during operation and has been used for the purpose of this risk assessment.
9
Holland-DAW_1-112-000 Page 12 Attachment Holland:DAW 1-112
Garrad Hassan America, Inc. Document : 41404/AR/02A
5.3 Individual risk
The results of the numerical modeling described in the Section 4.3 are shown in Table 5.2 below for an estimated 25 days of icing per year. The initiating probability is calculated according to WECO guidelines by estimating a constant rate of ice accretion along the whole length and tip face area of the turbine blades during periods of icing conditions. The typical range of ice thrown is taken to be the distance within which 90% of the ice throw or drop events would be expected to occur.
Throw Drop Ice fragment weight [kg] 0.5 1 0.5 1 Initiating probability [per year] 8400 4200 8400 4200 Typical Range [m] 0-150 0-160 0-45 0-44 Impact probability 90% Exceptional range [m] 150-290 160-320 45-97 44-81 Impact probability 10%
Table 5.2 Typical and exceptional ice throw and fall ranges
The all direction probabilities for ice throw and drop for 0.5 kg and 1 kg fragment weights considered is shown in Figure 5.1. These curves represent the probability of one ever-present 1m2 area being struck by an ice fragment in vicinity of Project site turbines assuming 25 days of icing per year.
10 1000g Throw 500g Throw 1 Avg Throw 1000g Drop 0.1 500g Drop Avg Drop 0.01
0.001
0.0001
Risk level per0.00001 square metre per year 0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 Distance from Turbine [m]
Figure 5.1 Ice fragment strikes estimated per m2 per year
10
Holland-DAW_1-112-000 Page 13 Attachment Holland:DAW 1-112
Garrad Hassan America, Inc. Document : 41404/AR/02A
Considering the risk of falling ice fragments from a stationary turbine, the level of risk by distance from the turbine and direction is presented in Figure 5.2.
0.00001
0.0001
0.001 N 320 0.01 NNW 280 NNE 240 0.1 200 1 WNW 160 ENE 120 80 40 W 0 E
WSW ESE
SSW SSE
S
Figure 5.2 Probability of ice fragment strikes per m2 by direction and distance
The results of the analysis indicate that the typical range of ice throw from the turbines is approximately 150 m, and the typical range (within 90% of time) of ice drop from the turbines is approximately 45 m. The results for the ice drop case indicate that the risk of a fragment of ice dropping and landing in a square meter a distance from the turbine drops sharply for distances beyond 60 m (in the range of the overhang of the wind turbine model).
The Figures 5.3 to 5.6 present the level of risk for the whole wind farm assuming the wind turbines will be operating during dangerous icing conditions.
11
Holland-DAW_1-112-000 Page 14 Attachment Holland:DAW 1-112
Garrad Hassan America, Inc. Document : 41404/AR/02A
Figure 5.3 Risk level of ice fragment strikes per m2 peryear–NorthpartoftheProject
12
Holland-DAW_1-112-000 Page 15 Attachment Holland:DAW 1-112
Garrad Hassan America, Inc. Document : 41404/AR/02A
Figure 5.4 Risk level of ice fragment strikes per m2 per year – North-central part of the Project
13
Holland-DAW_1-112-000 Page 16 Attachment Holland:DAW 1-112
Garrad Hassan America, Inc. Document : 41404/AR/02A
Figure 5.5 Risk level of ice fragment strikes per m2 per year – South-central part of the Project
14
Holland-DAW_1-112-000 Page 17 Attachment Holland:DAW 1-112
Garrad Hassan America, Inc. Document : 41404/AR/02A
Figure 5.6 Risk level of ice fragment strikes per m2 per year – South part of the Project
15
Holland-DAW_1-112-000 Page 18 Attachment Holland:DAW 1-112
Garrad Hassan America, Inc. Document : 41404/AR/02A
5.4 Control mitigation
Given the level of risk estimated within the parcels not under control of the Project from the wind turbines and that this estimate is based on several assumptions, it is prudent that a control method be employed at the Project to minimize the risk of potentially damaging ice fragments by implementing wind turbine control procedure when dangerous icing conditions are present.
The proposed procedures outlined in Section 4.3 should be sufficient to identify periods when icing is likely and to shutdown turbines when unsafe conditions are present. The ice detector and monitoring meteorological conditions also provides a direct measurement of the likelihood ice is starting to build up as well as the point at which icing conditions cease. It is important that all associated equipment for this system be diligently maintained and that the remote operator shutdown procedure is satisfactorily implemented by personnel so that all turbines will be shut down in the event of ice starting to build up. It is recognized that a risk may occur on start up of a turbine after a prolonged period of shutdown during icing conditions. In such circumstances, ice fragments may be released or thrown from blades in the first period of operation. This issue should be addressed by a suitable pre-startup inspection and remote startup procedure. With the proposed procedure and suitable pre-startup inspection and remote startup procedure, one can expect the ice build-up on the turbines to be no more than on any large stationary structure, with no risk of ice fragments being thrown from an operating rotor.
As with any large stationary structure, the risk remains of ice forming at a slow rate on the structure and dropping from the stationary turbine. By this method, when compared to an operating turbine, only a small amount of ice is likely to form. As this thaws, there will be some wind blow effect although that will be small on all but the lightest particles. GH estimates that only very high winds may cause fragments of any significant mass to fall or to be blown beyond 60 m of the turbine base. This is supported by the probability calculations presented in Figures 5.1 and 5.2.
As an additional safeguard, the Client will post warning signs along property lines and access ways to turbine locations.
16
Holland-DAW_1-112-000 Page 19 Attachment Holland:DAW 1-112
Garrad Hassan America, Inc. Document : 41404/AR/02A
6 CONCLUSIONS
GH has been contracted by the Client to undertake an assessment of the risk of ice fragments shed from wind turbines striking members of the public in the vicinity of the turbines at the Project.
It is concluded that if the proposed procedure and suitable pre-startup inspection and remote startup procedure, one can expect the ice build-up on the turbines to be no more than on any large stationary structure, with no risk of ice fragments being thrown from an operating rotor.
As with a large stationary structure, the risk remains of ice forming at a slow rate on the structure and dropping from the stationary turbine. By this method when compared to an operating turbine only a small amount of ice is likely to form. As this thaws, there will be some wind blow effect although that will be small on all but the lightest particles.
GH estimates that only very high winds in a specific direction may cause fragments of any significant mass to be blown beyond 60 m of the turbine base with a probability of fragment strike per square meter of approximately once in 65,000 years. Assuming 25 days of icing per year, this amounts to an individual risk for a stationary person present for all icing events located at 60 m of the turbine base of once in 10 years.
17
Holland-DAW_1-112-000 Page 20 Attachment Holland:DAW 1-112
Garrad Hassan America, Inc. Document : 41404/AR/02A
REFERENCES
1 C Morgan et al, “Wind energy production in cold climate (WECO)”, ETSU contractor’s report W/11/00452/REP, UK DTI, 1999.
2 C Morgan and E Bossanyi, “Wind turbine icing and public safety - a quantifiable risk?”, Proceedings of Boreas III Conference, Sariselka, Finland 1996.
3 E Bossanyi and C Morgan, “Wind turbine icing – its implications for public safety”, Proceedings of European Union Wind Energy Conference 1996.
4 C Morgan, E Bossanyi and H Seifert, “Assessment of safety risks arising from wind turbine icing”, Proceedings of EWEC ‘97 Conference, Dublin 1997.
5 C Morgan, “Assessment of Ice Throw Risk for the Proposed Huron Wind Farm”, GH report 3174/BR/01 Issue A, 30 April 2002.
18
Holland-DAW_1-112-000 Page 21 Attachment Holland:DAW 1-112
Garrad Hassan America, Inc. Document : 41404/AR/02A
LIST OF TABLES
2.1 Wind turbine parameters
5.2 Typical and exceptional ice throw and fall ranges
19
Holland-DAW_1-112-000 Page 22 Attachment Holland:DAW 1-112
Garrad Hassan America, Inc. Document : 41404/AR/02A
LIST OF FIGURES
2.1 Proposed locations and property line
3.1 Ice throw risk assessment procedure
4.1 Recorded ice throw data
4.2 Calculated probability of 1 kg ice fragment throw distances
4.3 Calculated probability of 1 kg ice fragment drop distances
5.1 Probability of ice fragment strikes per m2
5.2 Probability of ice fragment strikes per m2 by direction and distance
5.3 Risk level of ice fragment strikes per m2 per year – North part of the Project
5.4 Risk level of ice fragment strikes per m2 per year – North-central part of the Project
5.5 Risk level of ice fragment strikes per m2 per year – South-central part of the Project
5.6 Risk level of ice fragment strikes per m2 per year – South part of the Project
20
Holland-DAW_1-112-000 Page 23