Content uploaded by Claire Y. Barlow
Author content
All content in this area was uploaded by Claire Y. Barlow on Oct 23, 2019
Content may be subject to copyright.
Wind turbine blade waste in 2050
Pu Liu, Claire Y. Barlow
⇑
University of Cambridge Institute for Manufacturing, 17 Charles Babbage Road, Cambridge CB3 0FS, United Kingdom
article info
Article history:
Received 4 July 2016
Revised 23 January 2017
Accepted 4 February 2017
Available online 16 February 2017
Keywords:
Composites waste
Decommissioning
End-of-life
Wind turbine blade waste
abstract
Wind energy has developed rapidly over the last two decades to become one of the most promising and
economically viable sources of renewable energy. Although wind energy is claimed to provide clean
renewable energy without any emissions during operation, but it is only one side of the coin. The blades,
one of the most important components in the wind turbines, made with composite, are currently
regarded as unrecyclable. With the first wave of early commercial wind turbine installations now
approaching their end of life, the problem of blade disposal is just beginning to emerge as a significant
factor for the future. This paper is aimed at discovering the magnitude of the wind turbine blade waste
problem, looking not only at disposal but at all stages of a blade’s lifecycle. The first stage of the research,
the subject of this paper, is to accurately estimate present and future wind turbine blade waste inventory
using the most recent and most accurate data available. The result will provide a solid reference point to
help the industry and policy makers to understand the size of potential environmental problem and to
help to manage it better. This study starts by estimating the annual blade material usage with wind
energy installed capacity and average blade weight. The effect of other waste contributing factors in
the full lifecycle of wind turbine blades is then included, using industrial data from the manufacturing,
testing and in-service stages. The research indicates that there will be 43 million tonnes of blade waste
worldwide by 2050 with China possessing 40% of the waste, Europe 25%, the United States 16% and the
rest of the world 19%.
Crown Copyright Ó2017 Published by Elsevier Ltd. All rights reserved.
1. Introduction
Wind energy has become one of the most promising renewable
energy sources over the last two decades with the installed capac-
ity increasing from 7600 MW in 1998 to 364,270 MW in 2014
(GWEC, 2015). The capacity is expected to continue to increase,
although rates may vary in different geographical areas. The Global
Wind Energy Council (GWEC) predicts that the global annual
growth rate of wind power will exceed 12% between 2013 and
2018 (GWEC, 2014b). The European Wind Energy Association
(EWEA) predicts that by 2020 there will be 192 GW of wind capac-
ity supplying 14.9% of global electricity in 2020 (EWEA, 2014). The
International Energy Association (IEA) estimates that 15–18% of
global electricity will be produced from wind energy in 2050
(IEA, 2011). Despite these disparities, all the predictions indicate
that wind energy will continue to develop rapidly over the next
decade.
Although wind energy is often claimed to provide clean renew-
able energy without any emissions during operation (U.S.
Department of Energy, 2015), a detailed ecological study may indi-
cate otherwise even for this stage. The manufacture stage is
energy-intensive and is associated with a range of chemical usage
(Song et al., 2009). Disposal at end-of-life must also be considered
(Ortegon et al., 2012; Pickering, 2013; Job, 2014). A typical wind
turbine (WT) has a foundation, a tower, a nacelle and three blades.
The foundation is made from concrete; the tower is made from
steel or concrete; the nacelle is made mainly from steel and cop-
per; the blades are made from composite materials (Vestas,
2006; Tremeac and Meunier, 2009; Guezuraga et al., 2012). Consid-
ering these materials only, concrete and composites are the most
environmentally problematic at end-of-life, since there are cur-
rently no established industrial recycling routes for them
(Pimenta and Pinho, 2011; Job, 2013). Composite materials are
energy intensive to manufacture and some of the material has high
http://dx.doi.org/10.1016/j.wasman.2017.02.007
0956-053X/Crown Copyright Ó2017 Published by Elsevier Ltd. All rights reserved.
Abbreviations: AWEA, American Wind Energy Association; BoM, bill of materi-
als; CWEA, Chinese Wind Energy Association; EoL, end-of-life; EWEA, European
Wind Energy Association; GWEC, Global Wind Energy Council; IEA, International
Energy Association; kt, kilo tonnes; Mt, million tonnes; MW, mega watts; NREL,
United States National Renewable Energy Laboratory; O&M, operation and main-
tenance; PTC, Production Tax Credit for Renewable Energy; WT, wind turbine; WTB,
wind turbine blade.
⇑
Corresponding author.
E-mail addresses: pl384@cam.ac.uk (P. Liu), cyb1@cam.ac.uk (C.Y. Barlow).
Waste Management 62 (2017) 229–240
Contents lists available at ScienceDirect
Waste Management
journal homepage: www.elsevier.com/locate/wasman
value, which means they have strong recycling potential in terms
of both environmental and economic prospects (Shuaib et al.,
2015). This study focuses on the composite component of wind
turbine blades, looking at the waste inventory of all stages of their
lifecycle. Composites account for more than 90% of the weight of
WT blades (Liu and Barlow, 2016b). At present, most of the blades
are made from polymer composite reinforced with mainly glass
fibre, some carbon fibre and the hybrid combination of glass fibre
and carbon fibre (Collier and Ashwill, 2011). High-grade epoxy and
polyester are the mainstream resins used. Commonly adopted
manufacturing processes use Pre-impregnated fabric (Prepreg)
and Vacuum assisted resin transfer molding (VARTM) (Gurit
Composites, 2009). It is recognised that the materials and manu-
facturing techniques will evolve over time, but predictions vary.
Some predict that the proportion of carbon fibre will increase
(NEEDS, 2008; McKenna et al., 2016) and will lead to more serious
environmental impact from blade (Liu and Barlow, 2016a). How-
ever, current trends have provided no clear support for this trajec-
tory, so it may be that manufacturers are impeded by the high cost
of carbon fibre (Liu, 2016).
A few studies have been carried out on different aspects of the
ecology of wind energy. For the blade waste volume, Red estimates
there will be 260,000 tonnes material used to manufacture wind
turbine blades in 2008 and this number will increase to 1.18 mil-
lion tonnes in 2017 (Red, 2006). Albers notes that every one-
kilowatt of wind power needs ten kilograms of WT blade materials
(10 kg/kW or 10 t/MW), predicts that there will be nearly 50,000
tonnes of blade waste in 2020 and that this number will exceed
200,000 tonnes in 2034 (Albers, 2009). Andersen adopts Albers’
blade material demand figure of 10 t/MW and predicts that the
amount of blade material that will need to be recycled annually
is 400,000 tonnes between 2029 and 2033. It will increase to
800,000 tonnes per year by 2050 (Andersen et al., 2014). It is clear
that there will be a significant number of end-of-life WT blades
needing to be decommissioned over the next two decades. It
should be noted that the wind power industry has developed
rapidly in both scope and technology in this period (Sieros and
Chaviaropoulos, 2012; Siemens AG, 2014), which is not taken into
account by these previous studies. Liu and Barlow attempt to
tackle this issue but only provide general information about the
blade size increasing and lifecycle contributing factors (Liu and
Barlow, 2015). The more detailed analysis of the present study
includes such significant factors as the effect of increased turbine
size on blade masses, the variation between different geographical
regions, and consideration of waste generation over the whole life
cycle.
Presently, most WT blade waste is sent to landfill, but this is not
an environmentally benign solution, and indeed many European
Union countries have forbidden the landfilling of composite waste
(Pickering, 2006). Awareness of this issue is rising and has been
highlighted in recent wind power studies. Hayman raises the recy-
clability problem of wind turbine blades and Larsen summarises a
few possible recycling options for WT blades (Hayman et al., 2008;
Larsen, 2009). Both of them point out that the relatively short his-
tory of the WT industry and low production volumes lead to there
being no successful industrial scale WT blades recycling processes
that have yet been well-defined and established. Other studies also
explore possibilities for reusing the composite WT blades including
remanufacture and reuse as structural components in buildings,
bridges or artificial reefs (Asmatulu et al., 2013; Falavarjani,
2012). A few ideas have been proposed and have been trialed in
laboratories, but none of these has emerged as the industrial path
of choice for end-of-life WT blades either because of technical or
economical problems. At the moment, wind turbine blade manu-
facturers and governments lack detailed information about this
potential blade waste problem. They are aware that end-of-life
materials management needs to be addressed, and are keen to
know how serious a problem it will be and what options will be
available. A comprehensive answer is needed for this question,
including how much waste will be generated in the future, its envi-
ronmental impact, and the range of possible options for dealing
with the waste.
1.1. Research objective
This study aims answer the first part of the question above
which is trying to quantitatively and comprehensively understand
the life cycle waste inventory of WT blades using accurate and
state-of-art data. This paper provides a full evaluation of the mate-
rial flows associated with all stages of the lifecycle of WT blades,
estimated over a timeframe extending to 2050. Material is used
in the manufacture of the WT blades and during their service life,
to repair damage for example. At the end of their service life, the
blades are decommissioned and become end-of-life waste mate-
rial. The magnitudes of all these material usage and waste streams
are estimated using current global data and growth predictions
under different scenarios.
1.2. Paper structure
Research methodology and the logic behind the calculations are
introduced in Section 2. The blade material required per unit rated
power is analysed in Section 3.1 followed by the estimation of total
blade material usage presented in Section 3.2. Then the lifecycle
waste contributing factors from manufacture to end-of-life are dis-
cussed in Section 3.3. The waste inventory and model limitations
are presented in Sections 3.4 and 3.5 respectively. Finally, conclu-
sions are presented in Section 4.
2. Methods
The calculation starts from the manufacture stage. An estima-
tion of the amount of material used for WT blades globally requires
a statistical method with input from many different sources. We
need to know the amount of blade material required per unit wind
power, and to quantify how this changes over time with the evolu-
tion of turbine design and especially size.
The blade material usage is related to blade size and the blade
size is normally determined by its rated power. Generally, a high
rated power wind turbine needs large blades and this goes with
high blade material usage. Nevertheless, the relation between
blade size and rated power is only roughly proportional, not
directly proportional. In order to analyse the relation between
blade rated power and blade weight, we collect blade weight data
for 56 models produced by 14 wind turbine blade manufacturers
and divide them into five classes. In each class, the blade masses
are summed then divided by the sum of the turbine rated power
to obtain the average blade material required per unit rated power
(tonnes/MW) (Section 3.1).
The size of the wind power generation capability is then esti-
mated. Data on the current annual wind power installed capacity
and average rated power of new installed turbines is provided by
wind power associations, together with some predictions for the
future growth of the industry. These are used with the blade mate-
rial per unit rated power to calculate the total blade material
usage. For each specific year and region, we use the average rated
power in this region at this year to find the matched blade material
required per unit installed capacity (t/MW). We then use the unit
material requirement multiplied by the installed capacity (MW)
to get the total blade material usage (t) for this region during this
230 P. Liu, C.Y. Barlow / Waste Management 62 (2017) 229–240
period of time (Section 3.2). This blade material will become the
end-of-life (EoL) waste when the blades are decommissioned.
EoL waste does not constitute the full blade inventory. Wastes
arise from the whole blade lifecycle including the manufacture,
transportation, operation and maintenance and end-of-life. We
use the percentage of blade weight to represent waste levels, since
the amount of those wastes is proportional to the blade size (the
larger the blade, the more waste when the waste level is fixed).
Waste sources in addition to EoL waste are manufacturing in-
process waste, defective blades, testing blades, routine mainte-
nance, accidental damage and blade upgrading. Details are
explained in Section 3.3.
Finally, we sum the waste generated in each region and each
year to estimate the total amount of WT blade waste material that
will be generated over the period 2018–2050. Parts of this will
arise from manufacture of new blades and in-service waste, but
the picture will be increasingly dominated by the end-of-life waste
as WT blades are decommissioned (Section 3.4). The logic flow of
waste inventory estimation is presented in Fig. 1.
2.1. Data sources
The installed wind power capacity data are publicly available
from multiple wind energy associations. Blade specifications
including the model, weight, rated power and length are partially
publicly available from wind turbine specifications database web-
sites and blade manufacturers’ advertising materials; however this
has been augmented using 21 confidential bills of materials
received directly from wind turbine blade manufacturers through
site visits and interviews with technical directors. Data on manu-
facturing waste, operation and maintenance waste and end-of-
life waste have been collected through interviews with blade man-
ufacturers and wind farm operation and maintenance (O&M) ser-
vice providers and analysed by the researchers.
3. Results and discussion
3.1. Blade mass per unit rated power
The first step was to collect data on 56 mainstream wind tur-
bine blades (WTB) ranging in size from 500 kW to 8 MW, and orig-
inating mainly from US, Europe and Asia WTB manufacturers. The
blades are classified into the following size ranges: less than 1 MW,
1–1.5 MW, 1.5–2 MW, 2–5 MW and larger than 5 MW. There is a
continuing trend for wind turbines to up-scale, so usually the more
up-to-date turbines have higher rated power and larger blades. The
less than 1 MW class covers most of the early experimental tur-
bines and the early stage commercial turbines. The 1–1.5 MW,
1.5–2 MW and 2–5 MW classes cover most of the matured and
maturing commercial onshore wind turbines models and is also
projected to cover future onshore turbines for the next ten years.
The larger than 5 MW class is an offshore wind turbine class.
The finished blade masses from the different manufacturers are
presented in Fig. 2, as a function of the wind turbine rated power. A
Fig. 1. Logic flow of waste inventory estimation.
P. Liu, C.Y. Barlow / Waste Management 62 (2017) 229–240 231
clear trend linking blade size and power rating can be seen,
although there is quite a lot of variability in the data mainly
because the blades are manufactured at different times and
designed to be used in different wind speeds. An average value
of blade mass per unit rated power is needed for subsequent anal-
ysis, and is calculated for each of the turbine size class ranges. For
each turbine size class, the blade masses are summed then divided
by the sum of the turbine rated power to obtain the blade mass per
unit rated power (tonnes/MW). The results are presented in Fig. 3.
It can be seen here that the standard errors in all groups are lower
than 1 which shows that there is no extreme data in the sample
that has been selected. Additionally, as shown in Fig. 4, the blade
mass best fit polynomial curve is very close to the United States
National Renewable Energy Laboratory’s (NREL) prediction
(Fingersh et al., 2006) on the blade mass scaling curve. This indi-
cates that the blade mass sample is appropriate and representative.
The mass per unit power is lowest for the smallest wind tur-
bines, <1 MW, and it increases with the size of blades to reach
the highest value in large onshore blades, 2–5 MW. Simple geo-
metric arguments indicate that when the blade length is doubled,
the blade volume is increased by 2
3
, 8-fold. So for the same mate-
rial and same design, the blade mass would increase 8-fold. In fact,
as shown in Fig. 4, the blade mass does indeed increase with size,
but at a lower rate than predicted by the conventional mass scaling
law. As shown in Fig. 3, the blade mass per unit rated power of the
most up-to-date super-large offshore blade (>5 MW) is even
slightly lower than the large blade class (2–5 MW). These mass
reductions are due to developments in blade technology leading
to more efficient structural design, lower safety factors, lighter
materials and improved manufacturing techniques (Liu, 2015).
The results of blade mass per unit rated power (8–13.4 t/MW)
are similar to Henning’s results (10 t/MW) (Albers, 2009), but our
result have improved accuracy and have also considered the effect
of wind turbine upscaling.
3.2. Total wind turbine blade material usage
The amount of blade mass per unit rated power (t/MW) has
been calculated above. We need this number for the matched aver-
age rated power with the installed capacity (MW) to estimate the
total material usage for each region at any specific time. The aver-
age rated power for a single new installed turbine and annual
installed capacity depend on regional features.
Each region has its own strategy for developing wind power and
exhibits different features which will affect the blade waste level.
Europe, China, United States and rest of the world are selected as
0.00
5.00
10.00
15.00
20.00
25.00
30.00
35.00
40.00
0123456789
Single blade mass/tonnes
Rated Power/MW
Sinoma Gamesa Vestas Nordex Enercon DeWind Alstom
LZFRP GDUPC Mulbird Senvion Adwen Siemens LM Power
Fig. 2. Blade mass VS blade rated power. Modified from: (Liu and Barlow, 2015).
Fig. 3. Blade mass per unit rated power for the different turbine size classes.
232 P. Liu, C.Y. Barlow / Waste Management 62 (2017) 229–240
the four major wind energy markets based on their similar large
volume installed capacities. The data before 1998 is available for
Europe and United States only, but is not comprehensive for all
regions. The installed capacity before 1998 is very small compared
to the later installed capacity which means the effect on the final
results of the missing data is negligible. Hence, we decided to dis-
card data from before 1998 and consider only the waste levels after
1998. Note that the latest 2015 data is not yet available. Therefore,
the wind power installation historical data range was selected to
be from 1998 to 2014.
Based on information from American Wind Energy Association
(AWEA), Chinese Wind Energy Association (CWEA), European
Wind Energy Association (EWEA) and Global Wind Energy Council
(GWEC) (Anthony, 2014; GWEC, 2014a; EWEA, 2013; CWEA,
2014), as shown in Fig. 5, it can be seen that the commercial wind
energy industry started first in Europe where more than 50% of glo-
bal new wind turbines were installed between 1998 and 2006, and
the growth rate has continued steadily since then. The new
installed wind turbine sizes are increasing. The average rated
power of new installed wind turbines in Europe exceeded 1 MW
in 2000, 1.5 MW in 2006 and 2 MW in 2010 (Vitina et al., 2015;
IRENA, 2012; Woebbeking, 2012). Unites States also started devel-
oping wind energy early, installing 20% of the global new turbines
in 1999. In contrast to the stable European market, the US wind
market shows large fluctuations. The annual installed capacity is
strongly affected by the Production Tax Credit for Renewable
Energy (PTC) (Wiser and Bolinger, 2015). At the peak, US installed
177.6 GW wind energy in 2012, equivalent to 29% of the global
market share, but it then dropped severely to 14 GW in 2013. Its
average new installed wind turbine rated power exceeded 1 MW
in 2000, 1.5 MW in 2006 and 2 MW in 2015 (Wiser and Bolinger,
2015). Wind energy started late in China with only 617 MW wind
energy installed before 2005 (1.5% of Global installed capacity by
the end of 2004). Driven by a rapid increase in demand for electric-
ity and a strong renewable energy policy, China wind power then
experienced meteoric growth. The cumulative installed capacity
doubled every year during the period 2005–2009 and by 2010
China was the largest installed wind power capacity country. The
average new installed wind turbine rated power for China
exceeded 1 MW, 1.5 MW and 2 MW in 2007, 2010 and 2014
respectively (Liu, 2014). For the rest of the world, the installed
capacity has been steadily increasing since 2001. It is very hard
to find out the average new installed wind turbine rated power
for every single country. Hence we assume the average new
0
10
20
30
40
50
60
70
80
20 30 40 50 60 70 80 90
Blade mass/t
Blade length/m
Installed blade models NREL advanced blade mass scaling
Convenonal mass scalling Poly. (Installed blade models)
Fig. 4. Blade mass VS blade size. Modified from: (Liu and Barlow, 2015).
0
10000
20000
30000
40000
50000
60000
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Installed Capacity/MW
China United States Europe Rest of the World
Fig. 5. Annual installed capacity by region. Source: (Liu and Barlow, 2015).
P. Liu, C.Y. Barlow / Waste Management 62 (2017) 229–240 233
installed wind turbine rated power to be the same as the global
mean value from GWEC and to exceed 1 MW, 1.5 MW and 2 MW
in 2007, 2010 and 2014 respectively (GWEC, 2013, 2014a) (see
Table 1).
From above, the blade material usage is obtained from 1998 to
the end of 2014 based on the historical data. Installing capacity
prediction is required in order to find the blade material usage
for the future. Because the wind energy market is strongly affected
by energy policy and may show large fluctuations from year to
year, we decided to use the average of the last three years installed
capacity plus a growth rate predicted by the appropriate wind
energy association to estimate an annual installed capacity for
the year after the latest available data. For example, in order to
estimate the 2015 installed capacity we therefore average the
installed capacity of 2012, 2013 and 2014 and then multiply by
the predicted growth rate (100 + 14)%. Having established 2015
as the reference year, installed capacity for subsequent years is
estimated using only the predicted growth rate. The growth rate
affects future wind power installed capacity and the installed
capacity is the biggest factor determining the waste inventory, so
an accurate growth rate is crucial to this study. The future growth
rate is a prediction which is based on assumptions and is full of
uncertainties. Optimistic, normal and pessimistic scenarios are
commonly used to cover all the possibilities. Different wind energy
associations give different growth rate predictions. Normally local
energy associations are likely to provide more accurate growth rate
data than global predictions as the local energy association are
more familiar with local situations. We have attempted to find
growth rate predictions for each region, but only European and glo-
bal data has been identified. So we have used the growth rate for
Europe from the EWEA prediction, and have used the GWEC global
growth rate prediction for the other three regions.
Here, we adopt the same growth rate scenario settings as GWEC
did and the scenarios as ‘Base’, ‘Moderate’ and ‘Advanced’ from
GWEC are adopted for growth rate scenarios in this study. Three
scenarios are defined as follows: ‘‘The ‘Base’ scenario is based on
an assessment of current directions and intentions of both national
and international energy and climate policy, even though they may
not yet have been incorporated into formal decisions or enacted into
law. Examples of this would include the emissions reduction targets
adopted in Cancun in 2010, the various commitments to renewable
energy and efficiency at national and regional levels, and commit-
ments by governments in such fora as the G-8/G-20 and the Clean
Energy Ministerial. The ‘Moderate’ scenario has many of the same
characteristics as the Base scenario, taking into account all policy mea-
sures to support renewable energy either already enacted or in the
planning stages around the world, but at the same time assuming that
the commitments for emissions reductions agreed by governments at
Cancun will be implemented, although on the modest side. At the same
time it takes into account existing and planned national and regional
targets for the uptake of renewable energy in general and wind energy
in particular, and assumes that they are in fact met. The ‘Advanced’
scenario is the most ambitious, and indicates the extent to which the
wind industry could grow in a best case ‘wind energy vision’, but still
well within the capacity of the industry as it exists today and is likely
to grow in the future. It assumes an unambiguous commitment to
renewable energy in line with industry recommendations, the political
will to commit to appropriate policies and the political stamina to stick
with them. It also assumes that governments enact clear and effective
policies on carbon emission reductions in line with the now universally
agreed objective of keeping global mean temperature rise below 1.5–
2°C above pre-industrial temperatures.”(GWEC, 2014a).
By applying the growth rates to the historic installed capacity
data, the future installed capacity can be calculated. The historic
and future installed capacity form the full picture of installed data.
Next, we use the annual installed capacity (MW) multiplied by the
blade material required per unit power (t/MW), and from this the
total blade material usage in each year can be obtained. The result
is shown in Fig. 6.
3.3. Waste contributing factors
The total blade material usage calculated above is only a part of
the full blade waste inventory. Waste arises from the whole lifecy-
cle of wind turbine blades which comprises four stages: manufac-
turing, transport and installation, operation and maintenance, and
end-of-life. The blades themselves become waste at the end of
their service life, and are expected to form the largest fraction of
the total blade waste, but smaller amounts of waste arise in the
other stages in amounts that are proportionally related to the
amount of blade material (the materials actually present in the fin-
ished blade). For example, the amount of manufacturing in-process
waste is reported in terms of hundreds of kilograms per blade. We
can then represent the manufacturing in-process waste as a ratio
of the finished blade weight (%). We use the finished blade weight
as the reference value for material usage and multiply this by a
combined factor that includes all the other waste contributing
factors.
All of the contributing factors in the four stages affect the blade
waste, and the numbers may vary. For example, the in-process
waste is affected by the worker’s skills, the blade manufacturing
technology used and the manufacturing management practices at
the site. Hence, the waste contributing factors vary from region
to region, manufacturer to manufacturer and model to model. In
order to consider these variations, we set three scenarios for each
factor to give a better understanding of the full picture of blade
waste inventory. The ‘Central’ scenario is expected to be the clos-
est to reality and with the highest probability, i.e. the most likely
case. The ‘Low’ and ‘High’ scenarios represent the lowest and high-
est possible waste levels respectively.
Manufacturing in-process waste is estimated by subtracting the
mean finished blade mass from the bill of materials (BoM). The dif-
ference is the amount of material wasted during the manufacturing
process. The bill of materials contains the quantity and the types of
raw materials used in manufacture including the fibre fabric, resin,
structural adhesives, core, paint, metal accessories and manufac-
turing process consumable materials. It does not include working
protection consumables such as gloves, masks and containers
and packaging. Analysis of 21 BoMs provided by three blade man-
ufacturers for blades manufactured from glass fibre and epoxy
resin using VARTM technology revealed that the in-process waste
was between 12 and 30%, with median of 17%, of finished blade
Table 1
Average new installed single turbine capacity and blade mass per unit rated power.
Class Unit blade mass (t/MW) China US Europe Rest of world
Up to 1 MW 8.43 Pre 2006 Pre 1999 Pre 1999 Pre 2006
Between 1 MW and 1.5 MW 12.37 2007–2009 2000–2005 2000–2005 2007–2009
Between 1.5 MW and 2 MW 13.34 2010–2013 2006–2014 2006–2009 2010–2013
Between 2 MW and 5 MW 13.41 2014-Post 2015-Post 2010-Post 2014-Post
Larger or equal to 5 MW 12.58 Offshore Offshore Offshore Offshore
234 P. Liu, C.Y. Barlow / Waste Management 62 (2017) 229–240
mass (Liu, 2015). We assume that waste levels are comparable for
other manufacturers using the same manufacturing technology.
The other manufacturing technologies may bring different waste
levels. For example, the fibre usage of a 45-metre blade with
embedded bolts is 450 kg lower than the same model finished
using bolt hole drilling. Another example of variation is that Sie-
mens makes the blade in an unibody without structural glue (Inte-
gralBlade
Ò
technology). This technology improves the blade
integrity and is also able to reduce the blade weight and the waste
in polish and adhesives (Siemens AG, 2015). LM Power uses polye-
ster resin rather than the more commonly used epoxy resin and so
may have different resin usage level to other manufacturers (LM
Power, 2014). New direct infusion technology, used by some man-
ufacturers, can use a smaller pipe for resin transfer which could
reduce the resin residue waste (Bland, 2015).
The major in-process wastes are the dry fibre off-cuts, cured
composite off-cuts from the blade edge and root, resin residue in
flow mesh and container and the dust from the polishing process,
in proportions shown in Fig. 7.
Defects and testing blades are another two waste sources aris-
ing during the manufacturing stage. Defects are identified by
inspections at various times during the manufacturing process.
Small defects could be small regions with poor resin permeability
or slight bias in centre of gravity; such defects are quite common
and can be remedied during the manufacture stage. Defects requir-
ing discard of the whole blade or a whole blade component are
extremely rare and vary from model to model depending on the
maturity of the model. When new blade models are introduced
there may be high failure rates of this type, due to difficulties in
manufacturing techniques and the unfamiliarity of workers with
the new model. The rate of defects requiring discard of the whole
blade is typically around 0.05–0.2% (Liu, 2015). They are assumed
to be 0.05% for the low scenario, 0.1% for the central scenario and
0.2% for the high scenario in calculations.
Due to the certification requirements, a small number of blades
is made for mechanical testing purposes. For static tests they will
typically be loaded up to 150% of their designed loads for perform-
ing the stiffness and strength tests required for blade certification
and Finite Element Model validation (MTS Systems Corporation,
2012). New blade models also need fatigue testing involving the
automated cyclic loading of blades, typically at their resonant fre-
quency as a means of exciting the blade and achieving the desired
strain rate. Some of these static and dynamic testing blades cannot
be used in-service for electricity generation after the tests, and
hence are treated as testing blade waste, accounting for around
0.1% of all blades (Liu, 2015). The testing blade waste taken to be
0.05% for the low scenario, 0.1% for the central scenario and 0.2%
for the high scenario.
Some blades are damaged through improper hoisting, during
transport or during the installation process, but this rate is very
low (Liu, 2015). Waste generated in this stage is assumed to be
zero in this study.
Routine maintenance, accidental damage and blade upgrading
are the three major waste sources in the operation and mainte-
nance (O&M) stage for WTB. Routine maintenance includes clean-
ing, minor and major repairs. The repair of minor flaws or stone
damage is very common for most blades. Generally, 15 kg fibre,
resin and coating paint is enough for each of these minor repairs.
Minor defects may occur a few times during the blade lifetime
(Zhang, 2016). We assume they occur 2, 3 and 5 times for the
low, central and high scenarios respectively, which is equivalent
to 30 kg, 45 kg and 75 kg material consumption. Major repairs only
happen on specific blade batches and are usually caused by manu-
facturing defects or design defects. Such repairs typically involve
re-strengthening work on major structures such as shell bonding,
shear web bonding or the blade root. Each major repair job con-
sumes tens to hundreds of kilograms of fibre, resin and adhesives
(Zhang, 2016). In this study, the major repair material consump-
tion each time is taken to be 50 kg in the low scenario, 100 kg in
-
5,00,000
10,00,000
15,00,000
20,00,000
25,00,000
30,00,000
35,00,000
40,00,000
1998
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
2020
2022
2024
2026
2028
2030
2032
2034
2036
2038
2040
2042
2044
2046
2048
2050
Blade material usage/t
China United States Europe Rest of the World
Fig. 6. Annual WTB material usage. Data after 2014 is calculated based on the moderate growth rate scenario.
21%
27%
43%
9%
Dry fibre off-cuts
Cured composite off-cuts
Flow mesh with resin residue
Polishing dust
Fig. 7. Manufacturing in-process waste by weight.
P. Liu, C.Y. Barlow / Waste Management 62 (2017) 229–240 235
the central scenario and 150 kg in the high scenario. And the repair
demand rates for low, central and high scenarios are taken to be
5%, 10% and 20%. The total material consumption for minor and
major repairs is therefore equivalent to 0.5%, 0.9% and 1.6% of the
1.5 MW blade manufacturing material under our low, central and
high scenarios.
Quite a few blades break in accidents due to extreme weather: a
severe gust or high shear event can lead to loads that exceed the
blade design strength. Incorrect operation can also lead to exces-
sive loads on the blades and may considerably shorten the blade
life. Examples include incorrect shutdown sequencing, incorrect
pitch set or failure to maintain yaw alignment during high winds
(Malkin et al., 2015). A report indicates that those causes are
responsible for 1–3% of annual blade failures in the first ten years
of operation; the highest failure rates usually occur in the initial
five years (Malkin et al., 2015). Some failures need major repairs
and some of them require blade replacement. Such blades are trea-
ted as accidental O&M waste. The waste rate is 1%, 2% and 3% of
blade manufacturing material for the three scenarios.
Blade upgrading is another driver of waste during the operation
stage. With developments in blade aerodynamics, the newest
blades are able to capture more energy for the same wind turbine
compared to the blades made previously. The improved electricity
generating capacity means that some blades are replaced before
they reach the designed lifetime, which then leads to extra waste.
Some blade manufacturers also provide an aerodynamic efficiency
upgrading set which can be installed on blades to increase annual
energy production by 2–4% (AEP) (Siemens AG, 2014). Such blade
upgrade materials should be taken into account in the waste
inventory, but no information is available about the proportion of
blades upgraded and the amount of material involved. We assume
the upgrading waste is 2%, 5% and 10% of blade manufacturing
material for the low central and high scenarios respectively.
The wind turbine design lifespan is about 20 years. Currently,
there is no large scale commercial wind farm has that has yet
reached its design lifetime, so no one has experience about the
potential for wind turbine life extension. Gamesa presented their
research about the possibility of life extension at EWEA 2015
(Gamesa Corporación Tecnológica, 2015). They mentioned that life
extension for the tower and nacelle are relatively straightforward
but this is much more difficult for the blades. Their oldest blades
have been in operation for 17 or 18 years. Some them have already
suffered defects or fatigue problems at the shell bonding and root
connections which require major work to repair. Gamesa predict
that some blades could be used for more than 20 years, and maybe
up to 25 years, but it is not possible to extend the life to more than
27 years. Based on the above, we propose 18 years, 20 years and
25 years as the lifespan for blades under our three different
scenarios.
To summarise, the in-process waste and defective blade waste
are generated during the blade manufacturing process. Testing
blade waste is generated before volume production begins. The
time differences here are small, so we assume these three type of
waste are generated at the same time, which is the first year of
the lifetime of the blade. The routine O&M waste is generated by
the maintenance and repair which happen through the whole
blade lifetime, but generally small-scale repair and maintenance
work happens more frequently in the initial few years. The acci-
dental O&M waste is also mainly generated in the initial few years.
Hence, we assume all the O&M waste is generated in the sixth year.
The main purpose of blade upgrading is to improve the power gen-
eration efficiency. Blade upgrading is driven by relatively slow pro-
gress in aerodynamics research and blade technology. When
advances are made, it may take some time for the market to accept
change and respond. We assume that blade upgrading, with asso-
ciated waste generation, will not take place until the 16th year of
the lifetime of the blade. Based on the conclusions from Gamesa,
we conclude that a proportion of blades develop serious defects
and need major repair or to be decommissioned in the 17th/18th
year after commissioning (high scenario). Most blades have a
design lifetime of 20 years. These will be decommissioned in their
21st year (central scenario). As mentioned above, Gamesa also pre-
dicts that it will be possible to extend some blade lifetimes to
25 years without major defects arising. In this case, blades will
be decommissioned in their 26th year (low scenario) (Gamesa
Corporación Tecnológica, 2015).
All these waste contributing factors are summarised in Table 2
and the calculation logic is presented in Fig. 8. The combined fac-
tors for waste generated in the first three lifecycles stages, manu-
facturing, transport and installation, operation and maintenance,
are 15.6%, 25.1% and 45.0% for low, central and high scenarios
respectively.
3.4. Waste inventory
The blade waste inventory consists three types of waste: Man-
ufacturing waste, Service (O&M) waste and EoL waste. Manufac-
ture waste is the waste generated in manufacturing stage and
consists mainly of dry fibre offcuts, composite offcuts, resin residue
and vacuum consumables. Service waste is the material used dur-
ing the lifetime of the blade for routine maintenance, repair of acci-
dental damage and blade upgrading and is mostly fibre fabric and
resin. EoL waste refers to the retired blades, so mainly comprises
composite material (93%), with 2% PVC, 2% balsa and around 3%
metal, paint and putty (Liu and Barlow, 2016b).
The upper part of Fig. 9 presents the estimated global wind tur-
bine blade waste inventory in 2050 under different scenarios.
‘Growth rate’ is the predicted annual wind power installation
growth rate. The ‘affecting factor’ includes the waste contributing
factors during manufacture and the O&M stage. ‘Lifespan’ is the
wind turbine blade operation duration. Firstly, we aim to identify
the most likely waste weight in 2050. We therefore ascribe all
the variables to the most likely setting: the growth rate is set to
the moderate scenario and the waste affecting factors and lifespan
are set in the central scenario. This leads to an estimate of the most
likely blade waste weight of 43.4 Mt in 2050. An analysis is then
performed by looking at the ‘best’ and ‘worst’ cases. For the ‘best’
case, all the factors are chosen to benefit low waste volume such
as low manufacturing in-process waste rate, low new installed
capacity and long blade lifespan, giving a lower limit of blade
waste at 21.4 Mt. For the ‘worst’ case, factors are set in favour of
high waste volumes including the highest waste rates, high new
installed capacity and short lifespan, giving a blade waste upper
limit of 69.4 Mt.
The lower part of Fig. 9 presents the sensitivity analysis of vari-
ables. It shows the results variation in percentage (%) compared to
the most likely scenario as a benchmark. The growth rate is mainly
affected by the amount of wind turbine capacity installed, then the
number of blades manufactured and finally the blade material
usage. The higher the growth rate, the more of the newer models
of turbines installed, the larger will be the amount of waste in
the future. In the base scenario, the total waste will reduce 28%
compared to the benchmark. In the advanced scenario, it will
increase 19%. Affecting factors are related to the manufacturing
waste and O&M waste rates. With the high-level in-process waste
management and high quality blade (less repair required), the low
waste scenario will apply. In this case the total waste inventory is
14% less than the benchmark. By contrast for the high scenario, the
waste is 32% higher than benchmark. On the other hand, if the
blade service time is increased beyond the design lifetime, the
demand for new blades will be lower. The waste can be reduced
by up to 21% if the blades can serve for as much as 25 years. Con-
236 P. Liu, C.Y. Barlow / Waste Management 62 (2017) 229–240
versely, if the blade lifetime falls below the design lifetime, the
waste inventory may rise 10%. When all factors are considered,
the waste inventory in the lowest waste case is 51% lower than
the benchmark and the highest waste case is 60% above than
benchmark. There is a factor of up to 3.2 difference between the
best and worst scenarios, so there can be significant benefits from
advances such as improvement to the blade manufacturing tech-
nology to reduce in-process waste. Whatever scenarios are chosen,
however, the total waste will be a few tens of million tonnes in
2050 which will lead to serious environmental problems unless
proper solutions can be found.
We will now look at the waste types and the regional features.
In the following discussions, we use the most likely case (43.4 Mt),
moderate scenario for growth rate, and central scenario settings for
the other variables. As shown in Fig. 10, the annual scrap from
manufacturing and service steadily increases from 2018 with the
31.1
37.1 34.3
21.4
43.4 43.4 43.4 43.4
51.7
57.4
47.7
69.4
-
10.0
20.0
30.0
40.0
50.0
60.0
70.0
80.0
Growth Rate Affecng Factor Lifespan Upper/Lower Limits
Cumulave waste inventory
2050/Mt
-60%
-40%
-20%
0%
20%
40%
60%
80%
Growth Rate Affecng Factor Lifespan Upper/Lower Limits
Variaon compared to benchmark
Low waste level Benchmark (moderate) High waste level
Fig. 9. Upper: Global wind turbine blade waste 2050 in million tonnes (Mt), showing the effect of three different projection scenarios for each of three governing factors.
‘‘Affecting factor” includes waste contributing factors during both manufacture and O&M. The final column shows the maximum and minimum waste values obtained by
combining the factors. Lower: Waste variation compared to benchmark in %.
Table 2
Summary of waste contributing factors. Percentage represents % of finished blade mass.
Lifecycle Manufacturing Service End-of-life Total
In-process
waste
Testing blade
waste
Defective blade
waste
Routine O&M
waste
Accidental O&M
waste
Upgrading
waste
Year in which EoL
waste generated
Year of generation 1st 1st 1st 6th 6th 16th 18th–26th
Low scenario 12% 0.05% 0.05% 0.5% 1% 2% 26th 15.6%
Central scenario 17% 0.1% 0.1% 0.9% 2% 5% 21st 25.1%
High scenario 30% 0.2% 0.2% 1.6% 3% 10% 18th 45.0%
Blade material usage
Manufacture
• In-process waste
• Tesng blade
• Defecve blade
• Generate at 1st
year
• total 17.2% at
central scenario
O&M
• Roune service
waste
• Accidental damage
waste
• Generate at 6th
year
• 2.85% at central
scenario
Upgrading
• Generate at 16th
year
• 5% at central
scenario
End-of-Life
• 100% equivalent to
blade material
usage
• Generate at 21st
year in central
scenario
Blade waste
inventory @
125% in central
scenario
Fig. 8. Waste generation flow from manufacture to end-of-life.
P. Liu, C.Y. Barlow / Waste Management 62 (2017) 229–240 237
growth of new turbine installation. It reaches 500 kt in 2034 and
will keep increasing with the growth in blade manufacturing. By
contrast, end-of-life waste starts in 2018 under the central scenario
since the wind turbine installation data starts from 1998 and the
design lifetime is 20 years. It increases sharply to 500 kt per year
in 2029, overtaking the sum of all the other waste sources to form
the largest waste source at that time. This end-of-life waste stream
will annually generate more than 2 Mt blade waste in 2050.
The EoL waste in central scenario between 2020 and 2026 is
close to the data estimated by Albers (Albers, 2009). After 2027
our EoL waste data is higher since the up-to-date installed capacity
is adopted and the accuracy of blade material per MW installed is
improved. For the EoL waste between 2029 and 2033, our results
(around 500,000 t) are 20% higher than Andersen’s prediction
(400,000 t) (Andersen et al., 2014). This is because the unit blade
material demand during 2009 to 2013 is 12–13 t/MW which is
higher than Andersen’s 10 t/MW. The unit blade material demand
is more accurate in our research as it is directly calculated from
multiple real blade model weights rather than estimated from
more generic data. For 2050, Andersen estimates the blade waste
will exceed 800,000 tonnes per year. This figure assumes that the
cumulative installed capacity by 2030 will be 80 GW, and that
1/20 of this will be decommissioned by 2050. Our prediction is
based on a more detailed model which includes estimates for
annual changes in installation capacity.
The regional variations are illustrated in Fig. 11. China will need
to process 40% of the global blade waste; the equivalent figures for
Europe and United States are 25% and 16%. Since Europe started
installation of large scale wind farms earlier than other regions,
it will meet the end-of-life waste problem first. Two years from
now, there will be 15,000 tonnes of end-of-life blades needing to
be processed, increasing to more than 50,000 tonnes in 2022.
3.5. Model limitations
A number of assumptions and approximations have been care-
fully made in this work. We have used different scenarios to
demonstrate the sensitivity of the analysis to the various factors,
but the uncertainties in some of the predictions result in large
ranges in the estimates. The accuracy of results relies strongly
upon the input data availability and quality. A key uncertainty is
the wind energy growth rate prediction. Accurate regional growth
rate predictions are not available, so in this study we have used the
single figure of the global growth rate to provide estimates for the
growth in China, US and the rest of world. As the growth rates
strongly affect the total waste inventory, more accurate predictions
should be used in the analysis once they can be identified. The
other main area of interest is that we did not consider the effect
of transition to other manufacturing technologies such as unibody
manufacture technology because of lack of information: the bills of
-
5,00,000
10,00,000
15,00,000
20,00,000
25,00,000
2018 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038 2040 2042 2044 2046 2048 2050
TONNES
Manufacturing + Service End-of-life
Fig. 10. Global wind turbine blade waste projection up to 2050.
0
1,00,000
2,00,000
3,00,000
4,00,000
5,00,000
6,00,000
7,00,000
8,00,000
9,00,000
2018 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038 2040 2042 2044 2046 2048 2050
TONNES
China (40%) United States (16%) Europe (25%) Rest of World (19%)
Fig. 11. Regional wind turbine blade waste projection up to 2050.
238 P. Liu, C.Y. Barlow / Waste Management 62 (2017) 229–240
materials from manufacturers are classified (current data has been
gained through personal contact). Further information would be
required to investigate this aspect further. Another potential area
for refinement is that in the current work we have not included off-
shore (>5 MW) turbines in the final waste inventory estimation.
The reason is the current offshore installed capacity is much smal-
ler than onshore and the forecasts for future growth are very con-
fused. Most estimates, however, predict that offshore capacity will
not exceed 5% of the total wind energy market, so the effect of the
omission is expected to be limited. This could be reviewed when
further data becomes available.
4. Conclusion
This paper has systematically analysed and predicted the
amount of global wind turbine blade waste that will be produced
up to 2050 using the best available data from wind energy associ-
ations and blade manufacturers. Manufacturing waste, service
waste and end-of-life waste are the three major sources of blade
waste. Over the lifetime of the turbine, waste generated during
manufacturing and service adds between 16% and 45% of the mass
of the wind turbine blades. Sensitivity analysis of the contributing
factors reveals the most significant elements and provides insight
into where the wastes could be minimised. The balance between
the waste generated by the different contributing factors changes
over time. Manufacturing and service waste are currently the lar-
gest contributors, but end-of-life waste is increasing rapidly and
is projected to equal manufacturing and service waste in 2029.
The waste stream after this time is dominated by the end-of-life
blades which will become the biggest problem. The results show
that the end-of-life waste stream will annually generate more than
2 Mt in 2050 and cumulative blade waste in 2050 will lie between
21.4 Mt and 69.4 Mt with the most probable waste level being 43.4
Mt. Europe will face the problem first and ultimately China will
have the largest waste inventory.
Having quantified the amount of waste associated with wind
turbine blades, the next stage of the current research will be to
use the material flow data to estimate the environmental impact
of wind turbine blade manufacture and use in terms of CO
2
emis-
sions and energy consumption. Finally, end-of-life options for
decommissioning wind turbine blades will be explored with the
aim of providing environmentally favourable guidelines for
managing wind turbine blade waste.
Acknowledgments
The authors would like to thank the Industrial Sustainability
Research Group at the University of Cambridge and the industrial
cooperation partners for advice and support. This work was sup-
ported, in part, by China Scholarship Council (CSC). The authors
are also grateful to Jesus College, Cambridge for financial support.
References
Albers, H., 2009. Recycling of wind turbine rotor blades - fact or fiction? DEWI Mag.
34, 32–41.
Andersen, P.D. et al., 2014. Recycling of wind turbines. Available at: <http://
www.natlab.dtu.dk/english/Energy_Reports/DIER_2014>.
Anthony, J., 2014. AWEA U.S. Wind Energy Industry Market Update. Available at:
<http://www.awea.org/Resources/Content.aspx?ItemNumber=6386>.
Asmatulu, E., Twomey, J., Overcash, M., 2013. Recycling of fiber-reinforced
composites and direct structural composite recycling concept. J. Compos.
Mater. 48 (5), 593–608.
Bland, R., 2015. Manufacture of large composite structures by direct infusion
methods. In: Wind Turbine Blade Manufacture 2015. Applied Market
Information Ltd., Dusseldorf.
Collier, C., Ashwill, T., 2011. Materials and design methods look for the 100-m blade.
Windpower Engineering, May 2009. Available at: <http://www.
windpowerengineering.com/design/mechanical/materials-and-design-
methods-look-for-the-100-m-blade/>.
CWEA, 2014. 2013 China Wind Installations Statistics, Beijing.
EWEA, 2013. EWEA 2013 Annual Report Available at:
<http://books.google.com/books?hl=en&lr=&id=NkTWxqDX588C&oi=fnd&pg=
PA6&dq=EWEA+2013+Annual+Report&ots=nON53dqXv9&sig=
8v8mbJKvFGrAAb1arLScKo-6GDo> (accessed August 26, 2014).
EWEA, 2014. Wind Energy Scenarios for 2020, pp. 1–8. Available at: <http://www.
ewea.org/fileadmin/files/library/publications/scenarios/EWEA-Wind-energy-
scenarios-2020.pdf> (accessed November 12, 2014).
Falavarjani, B., 2012. Feasibility of using wind turbine blades structure as artificial
reef. EWEA, pp. 4–8. Available at: <http://proceedings.ewea.org/annual2012/
allfiles2/1553_EWEA2012presentation.pdf> (accessed October 2, 2014).
Fingersh, L., Hand, M., Laxson, A., 2006. Wind turbine design cost and scaling model.
Technical Report 29 (December), 1–43.
Gamesa Corporación Tecnológica, 2015. Gamesa life extension program
(unpublished presentation). In: European Wind Energy Association
Conference. EWEA, Paris.
Guezuraga, B., Zauner, R., Pölz, W., 2012. Life cycle assessment of two different
2 MW class wind turbines. Renew. Energy 37 (1), 37–44.
Gurit Composites, 2009. Wind Turbine Blade Structural Engineering Available at:
<http://www.gurit.com/files/documents/3_blade_structure.pdf>.
GWEC, 2013. Global Wind Statistics 2012, Brussels Available at: <http://www.gwec.
net/wp-content/uploads/2013/02/GWEC-PRstats-2012_english.pdf>.
GWEC, 2014a. Global Wind Energy Outlook 2014 Available at: <http://www.gwec.
net/wp-content/uploads/2014/04/Market-forecast-2014-2018.pdf>.
GWEC, 2014b. Market Forecast for 2014–2018, Brussels Available at: <http://www.
gwec.net/wp-content/uploads/2014/04/Market-forecast-2014-2018.pdf>.
GWEC, 2015. Global Wind Report Annual Market Update 2014 Available at:
<http://www.gwec.net/wp-content/uploads/2015/03/GWEC_Global_Wind_2014_
Report_LR.pdf>.
Hayman, B., Wedel-Heinen, J., Brøndsted, P., 2008. Materials challenges in present
and future wind energy. MRS Bull. 33 (4), 343–353.
IEA, 2011. Wind Energy Technology Roadmap. Springer-Verlag, Berlin/Heidelberg.
IRENA, 2012. Renewable Energy Technologies: Cost Analysis Series - Wind Power 1
(5), 4–35.
Job, S., 2013. Recycling glass fibre reinforced composites – history and progress.
Reinf. Plast. 57 (5), 19–23.
Job, S., 2014. Composite materials and end of life. High Value Manufacturing: Novel
Materials and Opportunities for the Circular Economy. Available at: <http://
www.green-alliance.org.uk/resources/Composite>. Materials and End of Life –
Stella Job.pdf.
Larsen, K., 2009. Recycling wind turbine blades. Renew. Energy Focus 9 (7), 70–73.
Liu, P., 2014. Wind Turbine Blade Market Trend. Interview with LZFRP CEO.
Liu, P., 2015. Confidential data collected through site visits to three China wind
turbine blade factories and private communication with blade manufacturer
technical departments.
Liu, P., 2016. Observations from wind energy exhibition and private communication
with blade manufacturer managers.
Liu, P., Barlow, C., 2015. An update for wind turbine blade waste inventory. In:
EWEA Annual Conference. EWEA, Paris.
Liu, P., Barlow, C., 2016a. The environmental impact of wind turbine blades. IOP
Conference Series: Materials Science and Engineering 139, 12032.
Liu, P., Barlow, C., 2016b. The environmental impact of Wind Turbine Blades
(Presentation). 37th Risoe International Symposium on Material Science,
1–16.
LM Power, 2014. Information from LM Power Employee on EWEA 2014. EWEA 2014
Annual Event.
Malkin, M., Byrne, A., Griffin, D., 2015. Does the Wind Industry have a Blade
Problem? Zackin Publications, Inc. Available at: <http://nawindpower.com/
online/issues/NAW1505/FEAT_02_Does-The-Wind-Industry-Have-A-Blade-
Problem.html>.
McKenna, R., Ostman, V.d., Leye, P., Fichtner, W., 2016. Key challenges and prospects
for large wind turbines. Renew. Sustain. Energy Rev. 53, 1212–1221.
MTS Systems Corporation, 2012. Wind Turbine Blade Testing Solutions Available at:
<http://www.mts.com/cs/groups/public/documents/library/dev_005130.pdf>.
NEEDS, 2008. New Energy Externalities Developments for Sustainability (NEEDS),
RS 1a: Life Cycle Approaches to Assess Emerging Energy Technologies. Final
Report on Off-Shore Wind Technology.
Ortegon, K., Nies, L.F., Sutherland, J.W., 2012. Preparing for end of service life of
wind turbines. J. Clean. Prod. 39, 191–199.
Pickering, S.J., 2006. Recycling technologies for thermoset composite materials—
current status. Compos. A Appl. Sci. Manuf. 37 (8), 1206–1215.
Pickering, S.J., 2013. Recycling and disposal of thermoset composites. In: Workshop
on Life Cycle Assessment (LCA) for Composites Gateway. Dartington Hall,
Devon, UK.
Pimenta, S., Pinho, S.T., 2011. Recycling carbon fibre reinforced polymers for
structural applications: technology review and market outlook. Waste
Management, New York, N.Y., vol. 31, no. 2, pp. 378–392.
Red, C., 2006. Wind turbine blades: big and getting bigger. Composite Technology.
Available at: <http://www.compositesworld.com/articles/wind-turbine-blades-
big-and-getting-bigger>.
Shuaib, N.A. et al., 2015. Resource efficiency and composite waste in UK supply
chain. Proc. CIRP 29, 662–667.
Siemens AG, 2014. Power Curve Upgrade Kits for Wind Turbine (unpublished
presentation). In: EWEA Annual Conference. EWEA, Barcelona.
P. Liu, C.Y. Barlow / Waste Management 62 (2017) 229–240 239
Siemens AG, 2015. Robust and low-weight: all-rounder blades. Siemens AG
Website. Available at: <http://www.siemens.com/global/en/home/markets/
wind/turbines/technology/blades.html> (accessed June 5, 2016).
Sieros, G., Chaviaropoulos, P., 2012. Upscaling wind turbines: theoretical and
practical aspects and their impact on the cost of energy. Wind Energy 15 (1), 3–17.
Song, Y.S., Youn, J.R., Gutowski, T.G., 2009. Life cycle energy analysis of fiber-
reinforced composites. Compos. A Appl. Sci. Manuf. 40 (8), 1257–1265.
Tremeac, B., Meunier, F., 2009. Life cycle analysis of 4.5 MW and 250 W wind
turbines. Renew. Sustain. Energy Rev. 13 (8), 2104–2110.
U.S. Department of Energy, 2015. Advantages and Challenges of Wind Energy. U.S.
Department of Energy’s Office of Energy Efficiency and Renewable Energy
(EERE) Website.
Vestas, 2006. Life Cycle Assessment of Offshore and on Shore Sited Wind Power
Plants Based on Vestas V90-3.0 MW Turbines.
Vitina, A. et al., 2015. IEA Wind Task 26: Wind Technology, Cost, and Performance
Trends in Denmark, Germany, Ireland, Norway, the European Union, and the
United States, 2007–2012, 54.
Wiser, R., Bolinger, M., 2015. 2014 Wind Technologies Market Report. Department
of Energy, U.S, p. 29.
Woebbeking, M., 2012. Turbine size: is big always beautiful? In: EWEA Annual
2012. Available at: <http://www.ewea.org/blog/2012/04/turbine-size-is-big-
always-beautiful/>.
Zhang, Z., 2016. Private Communication with the Technical Director of O&M Service
Provider (Kahn Wind). KahnWind.
240 P. Liu, C.Y. Barlow / Waste Management 62 (2017) 229–240
