Master Level Thesis Design and Assessment of a Large Commercial Photovoltaic

Master Level Thesis Design and Assessment of a Large Commercial Photovoltaic
Master Level Thesis
European Solar Engineering School
No.174, November 2013
Design and Assessment of a
Large Commercial Photovoltaic
System in Barbados
Master thesis 18 hp, 2013
Solar Energy Engineering
Student:
Terrence Haynes
Supervisors:
Frank Fiedler, Kaung Win Myat
Dalarna University
Energy and
Environmental
Technology
ii
Abstract
One of the main aims of this thesis is to design an optimized commercial Photovoltaic (PV)
system in Barbados from several variables such as racking type, module type and inverter type
based on practicality, technical performance as well as financial returns to the client. Detailed
simulations are done in PVSYST and financial models are used to compare different systems and
their viability. Once the preeminent system is determined from a financial and performance
perspective a detailed design is done using PVSYST and AutoCAD to design the most optimal
PV system for the customer. In doing so, suitable engineering drawings are generated which are
detailed enough for construction of the system. Detailed cost with quotes from relevant
manufacturers, suppliers and estimators become instrumental in determining Balance of System
Costs in addition to total project cost.
The final simulated system is suggested with a PV capacity of 425kW and an inverter output of
300kW resulting in an array oversizing of 1.42. The PV system has a weighted Performance Ratio
of 77%, a specific yield of 1467 kWh/kWp and a projected annual production of 624 MWh/yr.
This system is estimated to offset approximately 28% of Carlton’s electrical load annually. Over
the course of 20 years the PV system is projected to produce electricity at a cost of $0.201
USD/kWh which is significantly lower than the $0.35 USD/kWh paid to the utility at the time of
writing this thesis. Due to the high cost of electricity on the island, an attractive Feed-In-Tariff is
not necessary to warrant the installation of a commercial System which over a lifetime which
produces electricity at less than 60% of the cost to the user purchasing electricity from the utility.
A simple payback period of 5.4 years, a return on investment of 17 % without incentives, in
addition to an estimated diversion of 6840 barrels of oil or 2168 tonnes of CO2 further provides
compelling justification for the installation of a commercial Photovoltaic System not only on
Carlton A-1 Supermarket, but also island wide as well as regionally where most electricity supplies
are from imported fossil fuels.
iii
Acknowledgment
First and foremost, I would like to thank God Almighty for blessing me with the ability & opportunity to be
able to successfully complete this work. I am grateful to Hogskolan Dalarna & the Ministry of Education in
Barbados for affording me the opportunity of doing my Masters in Solar Energy Engineering at Dalarna
University. Many thanks go out to my family for their support and guidance throughout the course of this
endeavor. I would like to thank Tomilson Bynoe, Allen Benjamin and Frank Fiedler for the fruition of this novel
topic which is intended to assist in the development of Photovoltaics in Barbados. Without the help of Kaung
Myat Win I would still be stuck at square one in the design phase. He has been instrumental in providing design
guidance and recommendations throughout my thesis.
A huge thanks to Resco Energy Inc for all their support and allowing me to draw on their knowledgeable staff
and resources for shaping my thesis, without them, none of this would be possible. I would like to thank Martyn
Forde for being instrumental in me interning at Resco Energy and a special thanks to Shawn Speed, Fidel
Reijerse and Kevin Monsour who allowed me to intern at Resco and gain practical experience related to the
Engineering Procurement and Construction of Commercial Photovoltaic Systems. As painful as it was to come
in to work on weekends I would like to thank Shawn Speed for allowing me building access outside of work
hours to focus on my thesis which aided in the timely completion of my work. Many thanks to my Resco
supervisor David Booz for his continuous support and guidance in my PVSYST simulations and the overall
planning of the milestones for my thesis and ensuring that they were reasonable and that I achieved my targets
and that I covered all bases of design. Thanks to Michael Scott for design support and advice on inverter
manufacturers, racking and PVSYST simulations. A big thanks to Joe Ragno for overseeing & reviewing my
electrical design and electrical drawings. Thanks to Andrew Johnson for PVSYST direction and
recommendations. Thanks to Taylor Scott and Kyle Glover for their AutoCAD support along the way. A huge
thanks to Michael McGroary for his AutoCAD support and patience with helping me fine tune my engineering
drawings. I am thankful to Peter Stepien for assisting me in the estimating of my final project costs and final
balance of system pricing. Thanks to Shauna Haly for assisting me with the editing of this work.
I cannot fail to mention the help and support of Jason Hanschell who has supported me and encouraged me
from day one and has been instrumental in the overall organization and development of my thesis. Last but not
least I would like to thank all of my friends who have been instrumental in the successful completion of this
endeavor.
iv
Table of Contents
Abstract ............................................................................................................................................. iii
Acknowledgment ............................................................................................................................. iv
Table of Contents ............................................................................................................................. v
List of Figures .................................................................................................................................. vi
List of Tables................................................................................................................................... vii
Acronyms ........................................................................................................................................ viii
1
Introduction ........................................................................................................................... 1
1.1
Aims .............................................................................................................................................. 2
1.2
Method ......................................................................................................................................... 3
1.2.1
Preliminary Design Methodology .................................................................................... 3
1.2.2
Detailed Design Methodology .......................................................................................... 4
1.3
Previous work .............................................................................................................................. 4
2
Background Information .................................................................................................... 6
2.1
Boundary Conditions ................................................................................................................. 8
2.1.1
Grid Connection & Set up ................................................................................................ 8
2.2
Limitations ................................................................................................................................. 11
3
Preliminary Design ............................................................................................................ 11
3.1
Inverter Selection ...................................................................................................................... 12
3.2
Preliminary Module Selection ................................................................................................. 13
3.3
Preliminary Calculations .......................................................................................................... 15
3.3.1
String Sizing ....................................................................................................................... 15
3.3.2
String Numbering ............................................................................................................. 18
3.4
Preliminary Results ................................................................................................................... 19
3.4.1
Inverter Level Summary .................................................................................................. 19
3.4.2
System Level Summary .................................................................................................... 21
3.5
Preliminary Financial Analysis ................................................................................................ 22
3.6
Racking Selection ...................................................................................................................... 23
4
Detailed Design .................................................................................................................. 24
4.1
Feasibility & Site Visit .............................................................................................................. 25
4.2
Submission of Interconnection Agreement .......................................................................... 25
4.3
Structural Evaluation ................................................................................................................ 27
4.4
Shading Analysis........................................................................................................................ 27
4.5
DC Design ................................................................................................................................. 29
4.6
AC Design .................................................................................................................................. 29
4.7
BOS Takeoffs ............................................................................................................................ 30
4.8
Final System Simulation ........................................................................................................... 31
4.9
Final Simulation Results ........................................................................................................... 32
5
Results.................................................................................................................................... 33
5.1
Engineering Drawings .............................................................................................................. 33
5.1.1
Array Layout ...................................................................................................................... 33
5.1.2
Single Line Diagram ......................................................................................................... 35
5.1.3
Conductor Schedule ......................................................................................................... 35
5.1.4
Module Interconnection Drawing ................................................................................. 37
5.2
Project Costs .............................................................................................................................. 39
5.3
Financial Model ......................................................................................................................... 39
5.4
Proposed Incentives ................................................................................................................. 41
v
6
Discussion ............................................................................................................................ 42
7
Conclusion ............................................................................................................................ 44
8
References............................................................................................................................. 46
List of Appendices.......................................................................................................................... 47
Appendices ...................................................................................................................................... 48
List of Figures
Figure 1: Aerial View of Carlton Site (Bing, 2013) ................................................................................ 2
Figure 2 Energy breakdown of Small Island Development States (SIDS) (US Energy Information
Administration, 2010) ................................................................................................................................. 2
Figure 3 First Solar and Multicrystaline DC Power Output vs Temperature (Strevel et al. 2012) 5
Figure 4 3-D Render of Carlton Supermarket with conceptual solar panel design (Hanschell, 2013)
........................................................................................................................................................................ 7
Figure 5 Plan view of Carlton showing Section numbering and final module layout ..................... 7
Figure 6 BL&P Service Connection to Carlton A-1 (BL& P Company Limited, 2013) ................ 8
Figure 7 Carlton’s Monthly Electricity Usage from Dec 2011- Nov 2012 ....................................... 9
Figure 8 Contours or predicted 700 year return period peak gust wind speeds (mph) at a ......... 11
Figure 9 Top 10 PV suppliers of 2012 based on Megawatts shipped (Zipp, 2013) ......................14
Figure 10 Effect of variation on string length and module count per SMA 15kW inverter ......... 17
Figure 11 Array voltage & Power sizing for Trina 250W Maximum possible string length of 20 (Screen shot
from PVSYST) ...........................................................................................................................................18
Figure 12 Picture of Carlton Roof Section D detailing Roof Material ............................................. 24
Figure 13 Picture showing proposed mounting for Carlton’s trapezoidal roof (Snapnrack, 2013)24
Figure 14 Net Electricity Production to the Grid (BL&P, 2011)...................................................... 26
Figure 15 Total electricity production to the grid (BL&P, 2011) ....................................................... 26
Figure 16 Shading of Roof Section F of Carlton during mid-morning ............................................ 27
Figure 17 PVSYST Shading Model for Carlton A-1 ........................................................................... 28
Figure 18 Carlton A-1 shading loss after day of Simulation of Dec 21st 2013 ................................ 29
Figure 19 BOS Cost Breakdown 10MW, Fixed Tilt Multi c-Si Project in the US (Aboudi, 2011)30
Figure 20 Voltage drop calculation for Inverter #4 to AC Panel Board (Southwire Company, 2005)
...................................................................................................................................................................... 33
Figure 21 Cumulative Cash Flow of a 425kWp system over 20 years at a fixed cost of electricity of ($0.35/kWh)
...................................................................................................................................................................... 40
Figure 22 Cumulative Cash Flow of a 425kWp system over 20 years at a elecricity escalation rate of 3% annually
...................................................................................................................................................................... 41
Figure 23 List of existing PV installations in Barbados (SMA Portal, 2012) .................................... 48
vi
List of Tables
Table 1 Carlton Roof Sections showing irradiation on each section and losses with respect to optimum tilt and orientation
........................................................................................................................................................................ 8
Table 2 Minimum allowable PV System Size to meet Clients’ Requirements .................................................... 9
Table 3 Remuneration rates to the customer based on connection type to the grid ............................................... 10
Table 4 Module Characteristics for 4 module types being analysed .................................................................. 16
Table 5 Inverter losses for varying Trina 250W string lengths ........................................................................ 17
Table 6 Inverter Input Details for Trina-250 P05A module ......................................................................... 19
Table 7 PVSYST simulation results for different module brands per 15kW inverter ..................................... 19
Table 8 Array losses for different module brands per 15kW inverter .............................................................. 20
Table 9 System summary of various PV Systems based on module brands ...................................................... 21
Table 8 Cost table for various PV Systems based on module brand............................................................... 22
Table 9 Financial analysis for various system types ........................................................................................ 23
Table 12 Final System PVSYST Simulations ............................................................................................ 32
Table 13 Final pricing of the proposed 425KWp PV System for Carlton A-1 ............................................... 39
Table 17 Financial Analysis Summary for 425kWp PV System proposed for Carlton A-1 ......................... 40
Table 15 Detailed Bill of Material Cost for Carlton ....................................................................................... 55
Table 15 Direct Labor breakdown for Carlton A-1 ....................................................................................... 56
Table 16 Incidental Labour metrics for Carlton A-1 ...................................................................................... 56
Table 17 General Expenses for Carlton A-1 ................................................................................................ 56
vii
Acronyms
AC – Alternating Current
ASCE – American Society of Civil Engineers
BL&P- Barbados Light and Power
BOM – Bill of Materials
BOS- Balance of System
DC – Direct Current
FCA- Fuel Clause Adjustment
FIT- Feed In Tariff
GFD- Ground Fault Detection
GPS- Global Positions System
GRND- Ground
LDC- Local Distribution Company
MID- Module Interconnection Drawing
NEC- National Electric Code
NOCT- Nominal Operating Cell Temperature
NREL- National Renewable Energy Laboratory
OCPD- Over Current Protection Device
OLD- One Line Drawing
PPA- Power Purchase Agreement
PR- Performance Ratio
RGS- Renewable Generating System
ROI- Return on Investment
RTU-Roof Top Unit
SLD – Single Line Drawing
STC- Standard Test Conditions
viii
1 Introduction
Barbados arguably represents one of the most optimal places in the world for the utilization of Photovoltaics for
the generation of electricity. This claim is backed by the close proximity of Barbados to the equator with
coordinates 13.1594° N, 59.5300° W. This ideal location affords the island a high amount of solar radiation daily
with an average daily radiation of 5.7kWh/m2 (Rogers et al. 2011) and with over 3000 sunshine hours annually.
Other factors which increase the attractiveness of PV as a renewable energy solution in Barbados is the fact that
the island relies heavily on oil & natural gas imports from Trinidad & Tobago with very little of its oil produced
locally, only approximately 1000 barrels per day (Rogers et al. 2011). The expenditure of Barbados on oil
imports amounted to USD $220 Million in 2010 (Rogers et al. 2011) and recent claims suggest that this figure
was close to USD $300 Million in 2012 representing 25% of the island’s total import bill (Williams, 2013). Large
scale investment in PV for electricity can significantly reduce this expenditure and divert the amount of foreign
exchange lost to these oil imports. In addition, it is recognized that large scale commercial implementation of
PV will reduce the island’s carbon foot print and help curb its CO2 emissions.
There have recently been Government incentives implemented such as waived taxes on all PV related imports in
addition to tax rebates on installed PV systems as well as the Energy Rider program which is a provisional feedin-tariff program for renewable generators. These incentives will increase the attractiveness of investing in PV
and help bolster wide scale PV penetration on the island. Figure 2 shows the Renewable Energy penetration in
Barbados as of 2010 which is miniscule. This makes PV an exciting prospect for making Barbados more energy
independent as government has sent an aggressive target of having about 29 % of all electricity consumption to
be generated from renewable sources by 2029 (Williams, 2013).
The topic “Design and assessment of a large (>200kW) commercial Photovoltaic System in Barbados” came
about from a local business owner who owns a chain of supermarkets across the island. He expressed interest in
have a PV system installed on the newest supermarket; Carlton supermarket which was constructed in 2011. I
had the privilege of doing a site visit in January 2013. Carlton is not only a supermarket but a complex which
houses several other local business. The supermarket is located on the west of the island in the parish of St.
Michael. The Bing aerial view Figure 1 shows the site in relation to the island as well as the zoomed image
showing the various roof section.
The owner his interests in having a large percentages of his electricity costs offset by solar due to the recent
developments in incentives and agreements made by the government of Barbados and the local utility. This, in
addition to the client’s staggering monthly electricity cost served to spark interest in a commercial PV system.
The client’s current daily electricity loads are as high as 6.7 MWh/day.
Based on a search conducted of existing installations on the island which can be found in Appendix A, it can be
seen that commercial PV in Barbados is almost nonexistent and small scale, with only one installation on the
island above 100kWp. There hasn’t been in-depth local design at the moment regarding optimal PV system
Design which has looked at different PV technologies and how they compare from a cost and performance
standpoint. In addition, the fact that this area of commercial PV is undeveloped on the island provides a
compelling reason for this thesis. It is intended to show that large scale design can be economical, can be
optimized from a technological standpoint and is financially attractive even without incentives. This work is
hoped to provide a solid basis technologically, financially and logistically for commercial entities considering
installing PV systems on the island. If this project goes through to construction, it will represent a novel task and
the largest on the island to date.
1
Figure 1: Aerial View of Carlton Site (Bing, 2013)
Figure 2: Energy breakdown of Small Island Development States (SIDS) (US Energy Information Administration, 2010)
1.1 Aims
The main aim of this master thesis is to design a constructible PV system to optimally cover the roof top of
Carlton Supermarket in Barbados which will serve to help offset the user’s dependence on electricity from the
Barbados Light & Power. The optimal system design is intended to be achieved via 3 main objectives;
1) Preliminary Design which evaluates and selects the best solar components to be used in the final PV system
for Carlton based on performance and cost measures.
2) Engineering Design of a final constructible system. The scope of design for this thesis includes, PVSYST
simulations and shading analysis, a full system sizing with an array layout, electrical sizing and configuration as
2
well as site logistics. These tasks are largely reflected in the engineering drawings designed from scratch which
are included in the Appendix O-T and are explained in detail in the results section.
3) Another aim of this thesis is to demonstrate the financial attractiveness of the proposed PV system over its
lifetime. This is done by determining detailed cost of the final system and using commonly employed financial
metrics to show the project’s viability. It must be noted that structural evaluation is outside the scope of this
thesis.
Lastly, this thesis will consider unique factors applicable to the island such as high annual operating module
temperatures; accessibility to suppliers, high wind loads, the ability of the grid to handle the desired feed-in of
the PV generator and lastly any limitations which could act as barriers to the feasibility of such an endeavor.
1.2 Method
All design tasks will be done under the auspices of Resco Energy, an Engineering, Procurement, Construction
and Maintenance company (EPCM) based in Mississauga Canada. Resco will provide guidance with all stages of
the design and assessment of this project because of its potential for real life execution in the future.
From a methodology standpoint, the execution of this thesis can be broken down into two categories;
Preliminary Design and Detailed Engineered Design.
1.2.1 Preliminary Design Methodology
1. The first step in this process was the defining of the boundary conditions surrounding the project and in
so doing identify critical assumptions and potential limitations associated with the designing of Carlton’s
PV System. These Boundary Conditions were formulated by looking at relevant information such as
utility connection requirements, collection of customer electricity billing information, clearly identifying
the client’s system expectations, climatic factors such as temperatures and wind speeds which affect
system design and performance.
2. The scope of solar components to be evaluated was determined based on previous work relating to the
performance of different module technologies under elevated temperatures which is typical to tropical
regions like Barbados. Various module types and inverter types are explored and analyzed in this section.
The corresponding specification sheets for each of the proposed components is gathered, analyzed and
input into Microsoft Excel to facilitate the design process.
3. The next step was the design and string sizing of the proposed modules. This sizing and design took into
account the operating voltage characteristics of the inverter, the different current and voltage
characteristics of the different modules under Open circuit Voltage(Voc) conditions at the lowest
expected module temperature, Maximum power point voltage (Vmpp) at the highest expected module
temperature, short circuit current (Isc) and Maximum power point current (Impp). These factors were all
calculated in a spreadsheet in Microsoft Excel and were important in ensuring that the modules were
correctly sized to the inverter. Several PVSYST simulations were performed for each module type until
the ideal string length was achieved. This ideal string length was based on the minimization of array
losses from PVSYST and correspondingly maximizing the specific output and performance ratios of
these subsystems.
4. After the optimized string lengths were selected for each module type per inverter, the various PV
systems were designed based on available roof area as well as the modularity of the subsystems ie the
number of modules per inverter. The total number of modules based on module dimensions must be
divisible by the number of modules per inverter.
5. The next step in this design was the determination of each of the system costs. These costs were
determined via obtaining quotes from various sales representatives and distributors for the various
components as well as from the use of online pricing tools in addition to speaking with Resco’s
estimators. The details of these costs are depicted and discussed in the Preliminary Design Section.
3
6. 4 independent PVSYST simulations were then run for the various systems and the corresponding
performance metrics such as performance ratio, specific output and system losses were analyzed.
7. Following the PVSYST simulations were the financial analysis of each of the simulated systems using
Resco’s financial model which was adapted for the cost of electricity in Barbados as well as to show
metrics for simple payback period and Levelized Cost of Electricity (LCOE).
8. The final step of the preliminary design was the selection of the best PV system based on the PVYST
performance results and amount of load offset as well as the prospective financial returns of the systems.
This final system design then acted as a starting point or input for the detailed system design.
1.2.2 Detailed Design Methodology
1. The first task of the detailed design was to construct an accurate model of the building in PVSYST for a
shading analysis. The dimensions used to construct an accurate model of the building were taken from
existing site plans and elevations provided by the client. This step was crucial in determining whether it
was feasible to place modules on each roof section and where to avoid placing modules due to shading
in addition to providing a preliminary idea of how to string modules which would potentially be shaded.
2. The next task was the creation of an array layout in Autocad LT 2013. The existing site plan was
imported into Autocad and the dimensions of the Trina module were used to create a block and the roof
was then covered in modules taking into account module clip spacing, provision for walkways, shaded
areas and offsets from the roof edges.
3. The next step was to determine roof subsystem module counts on the seven different roof sections
labeled A-G. Based on these numbers in addition to pictures taken during the feasibility visit in
December 2012, inverter locations were determined for the 20 different inverters used.
4. Next was the determination of the AC cable lengths from each of the inverters to the electrical room
which will house the AC Combiner Panel that the inverter outputs feed into. Southwire Voltage drop
calculator was then used to determine the corresponding conductor sizes for each inverter based on the
maximum acceptable voltage drop of 1.55% as was defined in the boundary conditions.
5. Creation of the Module Interconnection Diagram (MID) in Autocad was the next detail design step.
This MID depicts which inverter, string and input each module on the roof belongs to. The basis of this
design was predicated on the shading of the strings, proximity of module to the nearest inverter and the
minimization of the homerun length (end of string back to inverter). The details of this task are further
explained in the design section.
6. Creation of the Single Line Drawing (SLD) in Autocad. This detailed task involved the depiction of how
the PV system flows electrically from the module level up to the electrical connection at Carlton’s
existing Main Electrical Panel. The SLD overcurrent protection devices such as breakers and fuses were
all sized according to the NEC code and are depicted in the SLD shown in Appendix S.
7. Creation of a Site Coordination Plan in Autocad which demonstrates the logistics of the proposed site
set up during construction.
8. Lastly, detailed cost of the project from start to finish was done using Resco’s estimating software. The
details and breakdown of these cost are included in Appendix F. A financial assessment of the final
system was done indicating common financial metrics in addition to LCOE and discounted payback
period.
1.3 Previous work
In executing the literature search, it was divided into two sections;
1) Internal which looked at previous ESES thesis of relevance as well as
2) External which looked at publicly available sources. The main thesis of interest & relevance was the “Design
of 2.2MW Solar PV roof in Spain” (Chen, 2007). This was of particular interest because of the similar nature of
looking at PV parking structures which is hoped to be explored in the future as well as similarities in climate. It
was interesting to see that the slope selected for the modules was 5o which not the ideal slope of 30o for highest
4
annual solar output for that region. It was understood that this selection was made based on the optimization
between the amount of losses incurred from suboptimal tilt, passive cleaning effect as well as increased wind
loads due to higher tilt. This factor will have to be taken into account when designing the parking lot PV in
Barbados which can experience hurricane force winds. Avoiding extra structures as indicated by Chen (2007)
will be replicated in the design in order to keep mounting and installation cost to a minimum. Another take away
from the parking lot PV installation in Spain will be to determine whether there are any specific building permits
required by Town & Country Planning for overhead structures (shelters) in Barbados.
The “The potential of PV installations in SIDS-an example in the island of Barbados” (Rogers et al. 2011)
highlights the potential for PV in Barbados by illustrating the islands dependence on fossil fuels and lack of PV
penetration as shown in Figure 2. This report provides promising results for PV in Barbados as an analysis of a
2kWp installation has a load factor of 23% (Rogers et al. 2011) which is quite considerable compared to other
locations worldwide. As highlighted, traditionally one of the barriers which exists is the lack of awareness of the
general public to the benefits of PV. This awareness is gradually increasing as the recently reelected Government
of Barbados has Renewable Energy as one of its main mandates for its tenure.
Because of relatively high temperatures year round, average daily ambient temperatures lie between 28-31oC, an
important factor for the selection of the system modules will be it corresponding temperature coefficient of
performance. Figure 3 and the white paper published by First Solar (Strevel et al. 2012) act to illustrate the
significantly higher output at high temperatures of First Solar`s CdTe thin film modules over multicrystaline
modules. As a result, thin film Cde, HIT and Polycrystalline modules will all be analyzed for output vs cost.
Figure 3 First Solar and Multicrystaline DC Power Output vs Temperature (Strevel et al. 2012)
The idea that the performance ratio of modules is dependent upon geographic location and that it differs
between module technologies is supported by Huld et al (2010) which maps the performance of fixed PV
installations across Europe for different module technologies. This study shows a correlation of decreasing
annual relative module efficiency with decreasing latitudes. This further acts to highlight the importance of the
analysis of performance of different module types under elevated temperatures which are typical of low latitude
regions such as Barbados (13º 10 North).
A search was done for commercial PV installations in the Caribbean using Google and Google scholar but it
was unyielding. However, Resco Energy specializes in the Engineering Procurement and Construction (EPC) of
5
commercial PV systems has done the full EPC on numerous PV systems installed across Supermarkets in
Ontario. As such relevant techniques and design steps have been adopted and followed where applicable to
Barbados. A search has been conducted on what systems have been installed on the island thus far, and these
results are illustrated in Appendix A. Out of the 28 existing installations on the island only one of these
installations exceed 100kW in size. This helps corroborate the lack of large scale PV on the island and illustrates
that PV is still in an infantile stage in Barbados. As a result, if this thesis project is pursued in reality, it will be
the largest installation on the island to date.
2 Background Information
•
•
•
•
•
•
•
•
•
Barbados is located at latitude 13º 10 North of the Equator and 59º 32´ West of Greenwich
Solar irradiation variance: 4.8Kwh/m2 (Dec) - 6.2kWh/m2(Apr). The daily average over the year is 5.7
kWh/m2
Barbados Light & Power is a privately owned utility providing a public service. Canadian company
Emera owns 79.7% of company shares(Barbados Light & Power Co Ltd, 2013).
BL&P has 3 generating stations island wide; Seawell, The Garrison & Spring Garden which total
approximately 240MW generating capacity. Supplies electricity for >100,000 customers island wide.
Consists of several natural gas turbines and one slow speed diesel generator (Barbados Light & Power
Co Ltd, 2013).
96% of Barbados’ electricity comes from diesel and fuel oil; 90%of which is imported. Expenditure on
fuel imports is in excess of USD $300,000,000 per year.
Voltage is generally transmitted at 24kV distributed at 11kV and utilized residentially at 115/200V
50Hz(Barbados Light & Power Co Ltd, 2013).
Barbados has a net peak demand of 188 MW.
Price of electricity to user is USD $0.35/kWh which includes the cost of generating electricity,VAT, the
utility connection fee and the Fuel Clause Adjustment which is tied to the fluctuating price of oil.
Carlton Supermarket located in Black Rock Barbados was constructed in 2011. It is not only a
supermarket but a complex which houses several local businesses. Figure 4 shows a conceptual 3-D
render of Carlton Supermarket with modules oriented in the potrait position installed acrossed all roof
sections.
6
Figure 4 3-D Render of Carlton Supermarket with conceptual solar panel design (Hanschell, 2013)
Figure 5 Plan view of Carlton showing Section numbering and final module layout
7
•
•
Total available roof area across the seven roof sections is 3,680m2. The roof material is galvanized steel
and is corrugated with a trapezoidal shape. A detail picutre of the roof detail is shown in the racking
selection section.
The seven roof sections are shown in Figure 5 with their respective tilts and azimuths depicted in Table
1. The building azimuth is -126o. Table 1, attained through PVSYST shows that it can be feasible to
place modules on the least favourably oriented roof section, Section C, with an annual irradiation on the
plane still at 1843 kWh/m2 from which a substanital amount of energy can still be harvested annually.
Table 1 Carlton Roof Sections showing irradiation on each section and losses with respect to optimum tilt and orientation
(PVSYST 5.59, 2012)
Roof Section
Tilt
Azimuth
A
B
C
D
E
F
G
15
15
15
15
5
5
5
54
-36
144
-126
-36
144
-126
Annual irradiation
on plane per m2
1948kWh
1961kWh
1843kWh
1859kWh
1965kWh
1925kWh
1930kWh
Loss with respect
to Optimum
0.9%
0.3%
6.2%
5.4%
0.0%
2.1%
1.8%
2.1 Boundary Conditions
2.1.1 Grid Connection & Set up
1. One of the first steps was to determine what the building grid voltage connection is and at what
frequency the grid operates. This is a fundamental for determining what inverter type best suits feeding
into the grid from Carlton. Generally, residential systems utilize a 115V/50Hz single phase connection
from BL&P. However, it was confirmed with the building engineer along with the building electrical
SLD that the connection is 3-phase 400/230V wye 4 wire connection. There is also a 700kVA backup
generator connected which an automatic switch which closes when the power from the grid is lost. This
is shown in the SLD in the engineering drawings section as well as in Appendix S.
Figure 6 BL&P Service Connection to Carlton A-1 (BL& P Company Limited, 2013)
8
1. At least 20% of customer’s daily load is to be offset. This stipulation was verbally communicated by the
client upon meeting with him to determine his expectations of an installed PV system. Figure 7 shows
the monthly loads from December 2011- November 2012. These were extracted from the client’s
electricity billing information. Based on the initial PVSYST simulation done for the Trina PA05 250W
module which is the selected module for this project, the corresponding specific yield 1492 kWh/kWp
was used to determine the minimum system size required to meet the customers’ expectations. This was
found to be 303 kWp as shown in Table 2.
Figure 7 Carlton’s Monthly Electricity Usage from Dec 2011- Nov 2012
Table 2 Minimum allowable PV System Size to meet Clients’ Requirements
Total Annual Electricity
Consumption
Minimum annual generation
required by PV system (20% load)
Approximate Specific Output of
Trina PA05-250W module
Minimum Allowable PV System
Size (DC)
2,259,130 kWh
451,826 kWh
1492 kWh/kWp
303 kWp
2. Cost of electricity is $0.35 USD/kWh
This Cost of electricity was determined from the BL&P electricity bill dated 2012-07-18 and includes
VAT and the customer connection fee as well. This price is shown in Table which shows remuneration
to the client from the PV system based on the type of metering connection.
3. FIT at the time of writing this thesis $0.36 USD/kWh
Based on the BL&P Renewable Energy Rider (2012) the feed in tariff is determined as all kWh supplied
to the grid at 1.8 times the Fuel Clause Adjustment or 31.5 cents BDS/kWh whereby the FCA is
$0.40/kWh as taken from the BL&P electricity bill dated 2012-07-18.
9
Table 3 Remuneration rates to the customer based on connection type to the grid
CASE
All to grid
Excess to the grid
FIT
Cost of Electricity
BBD/kWh
0.72
0.71
USD/kWh
0.36
0.35
4. Assumed fixed exchange rate of $2.00BDS to $1.00USD for all conversions. This exchange rate has
been fixed as opposed to floating since the mid 1970’s.
5. Maximum allowable DC Voltage drop from the string home run to Inverter was 0.5%. Maximum
allowable AC voltage drop from the inverter to AC Combiner Panel is 1.55%. Maximum allowable AC
voltage drop from AC Panel Board to Building Main Panel is 0.5%
6. All costs are given in USD unless otherwise specified.
7. A crucial step in designing a PV system is the determination of the maximum and minimum operating
cell temperatures which act to influence the maximum and minimum strings lengths due to the module
temperature coefficient of voltage characteristic. The maximum operating cell temperature was assumed
to be 85oC based on the maximum ambient temperature recorded in Barbados in 2012 being 32◦C in
addition to the fact that the maximum ambient temperature recorded in the past 20 years not exceeding
36oC. The minimum assumed module operating temperature is 18oC. The graphs supporting these upper
and lower temperature limits are both shown in Appendix D. It is also assumed that the usual operating
cell temperature would be 60◦C. These temperatures were used as input parameters in PVSYST.
8. Racking system Designed Wind load Spans, rail weight and max cantilever are selected to withstand
130mph winds. It is assumed that structural analysis will also be done based on this wind speed.This is
based on the Caribbean application to The Standard ASCE 7 (the American Society of Civil Engineers)
“Minimum Design Loads for Buildings and Other Structures” (Gibbs, 2008). The following excerpt
taken from this standard justifies the usage of 130mph wind speeds: “Here, the basic wind speed is the
700-year return period wind speed divided by sqrt (1.6) which yields a design wind speed that is
consistent with the intent of the developers of the ASCE 7 wind speed map.” (Gibbs, 2008). This
contour map for the Caribbean is shown in Figure 8. The 700 year return wind speed for Barbados is
152 mph (Gibbs, 2008).Therefore the basic design wind speed used in wind loading calculations for
Barbados would be 152/sqrt(1.6) which is 120mph. Thus designing the PV system based on 130mph
wind speeds is conservative and bounds calculations done at 120mph.
10
Figure 8 Contours or predicted 700 year return period peak gust wind speeds (mph) at a
height of 10m in flat open terrain (ASCE 7 Exposure C) (Gibbs, 2008)
9. When available, the quoted shipping cost is included, otherwise shipping costs are assumed
conservatively to be 5% of the total cost of merchandise. All equipment is exempted from duty per the
Renewable Energy Incentives provided by the Government of Barbados.
10. Assumed 100 % client financed with no debt or equity.
11. VAT credited to the client if exempted.
2.2 Limitations
1. BL&P connection restrictions and size limitations.
At the time of writing this thesis the maximum connection size for grid installed renewable energy
generators is 150kW with a contractual agreement for a pilot period of 2 years whereby after this
duration the terms of agreement will be reevaluated by the Utility and then renegotiated by the FTC.
2. Structural report limitations and proposed layout & loading are to go to engineers at time of being
awarded the contract as the costs for this assessment cannot be incurred otherwise.
3. All prices for modules, inverters and other BOS components are given based on Q3 of 2013 pricing. It
is understood that this pricing can vary considerably depending on market conditions from Q3 of 2013
to Q2 of 2014 when this equipment is likely to be procured compared to price variation at Q2 of 2014.
4. PVSYST 5.59 is limited to the modeling of two different orientations. Because of this limitation, the
seven roof sections with different tilts and orientations have to be modeled in several simulations
independently and then integrated to a final comprehensive system report.
3 Preliminary Design
This section is meant to evaluate and explore the different merits of various PV components based on their
price, bankability, performance and technical feasibility. This section serves as a harbinger for the selection of
the final optimized system based on financial merits which take both cost and performance into account. The
final optimized system is then fully designed in the detailed design section.
11
3.1 Inverter Selection
The first major task was selecting an inverter type for Carlton. Due to the building having seven different roof
sections with different tilts and orientations, it was out of the question to have one central inverter. Too many
losses would have been incurred due to string output mismatch. Some thought was given to having a few central
inverters for different sections but this proved to be a seemingly unnecessary design challenge. Some
consideration was also given to micro inverters but for a system with 2000 plus modules this presented too
many points for system failure. In addition, micro inverters cost per watt ends up being substantially more than
for string or central inverters. The only seemingly appropriate solution which remained was the use of string
inverters. For commercial applications there are a number of compelling reasons to use string inverters which
are listed below (Elerchiman, 2013):
1. Optimized Inverter to Array Ratio/ Design Flexibility
In general central inverters have larger power capacity increments than string inverters & as a result poorer
modularity than string inverters. Also string inverters allow for greater design flexibility when dealing with
multiple tilt angles and orientations which central inverters are not capable of without significant losses.
2. Decreased DC BOS Costs
Central inverters require DC combiner boxes with aggregated string numbers with high numbers of strings
feeding into these boxes resulting in higher currents on the DC side being run into the inverter. This increased
ampacity requires large TECK cable to carry these large currents. This TECK cable is expensive on the order of
$50 USD/m. In addition to this, the cost of combiner boxes adds up for large systems. String inverters reduce
the use of this expensive cables on the DC side as well as avoids the use of combiner boxes. Another advantage
of these string inverters is that the fault current is limited on the DC side resulting is smaller OCPD.
3. Better Granular Monitoring
By using string inverters, the granularity of system monitoring is better and easier because of the increased
number of monitoring points. For central inverters to achieve such detailed monitoring, combiner string level
monitoring or zone level monitoring will be required which comes with an additional cost.
4. Reduction in Space of Infrastructure
For applications where space is limited such as carports, string inverters are the ideal solution as they are light
weight and can be wall mounted. In addition, by having string inverters, the cost of pouring a concrete pad is
avoided as well as the use of a heavy crane for installation.
5. Increases System Availability
This is greatly improved in a string inverter system. For example, in the case of an inverter failure in a
commercial system, only a small fraction of system production is lost, however in a commercial system where
there is only one central inverter, the system production is 100% reduced. This total lost in production with
central inverters can have a substantial effect in lost revenue as more time elapses before the inverter is repaired
/replaced.
6. Ease of Replacement
Having standardized string inverters as is the case for Carlton, allows for easy replacement when an inverter
goes down by keeping a spare on hand which can be swapped out with relative ease. This is a downfall with
Central inverters as a technician will have to be called to site to fix the inverter and if it requires replacement
logistics can be difficult with coordinating crane and personnel to install the new inverter. In addition there is
generally long lead time for procurement of the larger inverters which would have to be shipped from the
nearest distribution center. With string inverters, this step of repair will be done in parallel to having a new
inverter installed cutting back on down time, lost revenue and costs incurred by the client.
Once it was determined that string inverters would be used, the next step was to determine the best type for
Carlton. Given the grid requirements, the ideal inverter would be 3-phase, 50Hz with a 230/400V output and
with a large input Vmpp range for increased design flexibility. In addition, an inverter with more than one Mpp
tracker put input is ideal as it allows for flexible string design and allocation due to shading which could severely
reduce inverter output. Lastly, a high overall efficiency was desirous as well for optimal performance and
12
minimal losses. After reviewing several candidates such as KACO, Power One and SMA the best fit for this
application was deemed to be SMA TriPower for several reasons;
1) SMA as a brand is the number one source of inverters used on the island as can be seen from Appendix
A.Therefore potential clients will likely to be more inclined in accepting a brand that has been heard of before
on the island.
2) Bankability is another motive for selecting SMA as an inverter supplier as they are one of the most established
solar inverter manufacturers globally with over 25 Giga-Watts installed worldwide (Hermes, 2013). This track
record serves to attract investment, designer and client approval for the use of SMA inverters.
3) SMA TriPower 15000TL has a large Vmpp range of 360-800V which allows for a large design flexibility
allowing for use of different modules with varied voltage characteristics in addition to allowing designs of varied
string lengths. Another attractive feature of the SMA Tripower series is that it has an active Maximum Power
Point Tracking system called the SMA Opti Trac Global Peak. The advantage of this MPP tracking system is
that it can detect the presence of multiple maximum power point peaks due to shading and adjust to operate at
the optimal peak thereby minimizing shading effect. The section from the Tripower installation manual
describing this functionality is included in Appendix J. Lastly, with an overall efficiency of 98.5 %, this makes
the SMA TriPower 15000TL the ideal candidate for the design.
3.2 Preliminary Module Selection
With a plethora of PV module manufacturers on the market all of which have different models, the module
selection for PV designers and installers has never been greater. Because of this, some basic criteria needed to be
developed to base selection decisions on in order to refine the list of modules to be analyzed.
In the case of commercial installation such as Carlton, one of the main constraints for module selection has to
be made from an investment perspective of the customer, in other words, the bankability of a PV module
manufacturer is crucial in these commercial situations. The bankability of any investment is the degree of how
likely this investment will bring financial success and consequently profit. This allows for greater ease of
financing and equity investing for these million dollar projects. This means that well established manufacturers
are preferred, which can back their module guarantees over the 20 year life of the solar project despite of varying
economic conditions. In addition to a reputable name, clients will be looking for competitively priced modules.
They are not as concerned with the technological performance as most designers or engineers are. Figure 9
shows the most successful module manufacturers from 2012 in terms of Megawatts sold.
13
Figure 9 Top 10 PV suppliers of 2012 based on Megawatts shipped (Zipp, 2013)
A major consideration from a designer’s perspective will be the selection of an appropriate module &
technology based on performance of these modules in the specific climate for which they will be installed. For
instance, if modules are to be installed in an often overcast location, engineers and designers may pay more
attention to the selection of modules based on performance in low light conditions as opposed to any other
factor. In particular, for this thesis the performance of modules under elevated cell temperatures factors in quite
importantly in the selection process. Amorphous silicon modules, despite having a low temperature coefficient
of performance were not considered for this application due to their significantly lower module efficiency which
could not be justified at Carlton where the general design goal is to maximize system output while minimizing
cost.
Monocrystaline modules were not considered due to the ever decreasing efficiency gap between polycrystalline
and monocrystaline modules. The increase in cost of monocrystaline modules over polycyrstaline (in some cases
up to $0.10 more) is not justified for a marginal increase in module efficiency. Moreover, when monocrystaline
was modeled in preliminary assessments for this same site, the monocrystaline modules failed to outperform
multicrystaline modules when cost was taken into account.
After carefully reviewing costing information, speaking to relevant suppliers and distributors as well as factoring
in bankability and module performances under elevated temperatures, the following modules were chosen to
determine which ultimately would prove to be the most optimal module from a financial and performance
perspective:
1. Ying Li Green Energy 72 Cell YL280-35b
This module was chosen due to the fact that it is a 72 cell 280W module, the intent being to determine whether
a larger and higher power module would have a significant increase in output when subjected to a grid tied
installation. In addition, the bankability of Ying Li is high as it has been one of the top suppliers of modules
worldwide for years.
2. Trina P05A-250
In addition to its popularity as shown in Figure 9, this 250W module was chosen because of its relatively low
temperature coefficient of voltage for polycrystalline modules (0.32 %/◦C). Furthermore, this module was
14
competitively priced and it is one of the module types that RESCO has used successfully across many of their
installations. Lastly, there is a Trina supplier in Puerto Rico whose proximity to Barbados allows for better
logistics for shipping and procurement.
3. Sanyo HIT 240HDE4
Selection of this module was based on its high module efficiency (19%) and its marked increase in performance
under high operating temperatures relative to polycrystalline modules. Accordingly, this was the highest price
module ($1.35 USD/Wp) out of the four selected. The merit of the increased cost for increased performance is
explored below.
4. First Solar FS-390 CdTe
Lastly, First Solar’s FS-390 CdTe 90W module was considered for Carlton because of its superior performance
under high temperatures despite CdTe’s traditionally lower module efficiency compared to polycrystaline.
3.3 Preliminary Calculations
3.3.1 String Sizing
A necessary step in the process of calculating string lengths was defining the maximum allowable DC voltage for
a “low voltage” Generator based on the NEC 2011 Code which is adhered to in Barbados. This was determined
to be 600V for residential applications per NEC 2011 Code Section 690.7B which is also listed in Appendix B
NEC Code However, Carlton being a commercial site is allowed a DC system voltage of up to 1000V based on
the NEC 2011 code. Initially however all string lengths for modules were determined with this 600V upper limit
and consequently simulated in PVYSYST. All PVSYST Simulations were run for optimizing string lengths based
on the maximum inverter Vmpp voltage of 800V.
Table 4 shows the STC characteristics of each of the 4 module types being compared in this thesis. Also shown
in this table are the maximum and minimum voltage ranges throughout which the modules are expected to
output with the lowest output being at Vmpp at 85oC the hottest expected module temperature and the highest
expected module voltage of Voc at 18oC the lowest expected module temperature. Using Microsoft Excel and the
corresponding temperature coefficient of voltage for each module type Vocmin(at 18oC), Vmppmin (at 18oC) &
Vmppmax(at 85oC) was determined as recommended by PV for Professionals (Durscher et al. 2007). Once this is
done, the maximum and minimum module operating voltages coupled with the Vmpp voltage window of the
inverter (360V-800V) allow for a maximum and minimum possible string length for each module type as shown
in Table 4. The optimized string length for each module type is chosen by an iterative PVSYST process as
mentioned below.
15
Table 4 Module Characteristics for 4 module types being analysed
SANYO HIT-240HDE4
TRINA-250 P05A
Module Dimensions etc
Module Dimensions etc
Length(m)
Width(m)
Height(m)
Weight(kg)
1.61
0.861
0.035
16.5
Length(m)
Width(m)
Height(m)
Weight(kg)
1.65
0.99
0.038
19.5
V mpp(V)
35.5
V mpp(V)
Impp(A)
5.51
Impp(A)
8.01
V oc(V)
43.6
V oc(V)
37.8
Module Characteristics
Isc(A)
Module Characteristics
30.5
7.37
Isc(A)
0.109
V temp Coefficient (V/ C)
V oc(V) at 18 C
44.36
V oc(V) at 18 C
38.65
o
36.26
V mpp(V) at 18 C
o
31.18
o
28.96
V mpp(V) at 85 C
o
24.64
o
V temp Coefficient (V/ C)
o
V mpp(V) at 18 C
V mpp(V) at 85 C
Min String length inorder to operate
in Mpp range (360V)
Max string length to be within
Inverter V mpp
Chosen String Length
12
18.03
15
8.5
o
o
Min String length inorder to operate in
Mpp range (360V)
Max string length to be within Inverter
V mpp
Chosen String Length
Ying Li YL280-35b
First Solar FS-390
Module Dimensions etc
Module Dimensions etc
Length(m)
Width(m)
Height(m)
Weight(kg)
0.32
15
20.7
17
1.97
0.99
0.05
26.8
Length(m)
Width(m)
Height(mm)
Weight(kg)
1.2
0.6
6.8
12
V mpp(V)
35.5
V mpp(V)
Impp(A)
7.89
Impp(A)
1.9
V oc(V)
45
V oc(V)
60.5
8.35
Isc(A)
0.37
V temp Coefficient (V/ C)
V oc(V) at 18 C
46.17
V oc(V) at 18 C
61.35
o
36.42
V mpp(V) at 18 C
o
48.06
o
27.62
V mpp(V) at 85 C
Min String length inorder to operate in
Mpp range (360V)
Max string length to be within Inverter
V mpp
Chosen String Length
o
41.71
Module Characteristics
Isc(A)
o
V temp Coefficient (V/ C)
o
V mpp(V) at 18 C
V mpp(V) at 85 C
Min String length inorder to operate
in Mpp range (360V)
Max string length to be within
Inverter V mpp
Chosen String Length
Module Characteristics
13
17
15
47.4
2.06
o
o
0.2
9
13
10
The string lengths for each module type were then narrowed down and all optimized using an iterative process
in PVYST by varying string length between minimum allowable string length and maximum allowable string
length (shown in Table 4) while observing the corresponding effect on Perfornance Ratio, Specific Output as
well as the magnitude of inverter overload loss. Figure 10 shows an iteration done for TRINA 250W module
with varying string lengths showing performance ratio and specific production.
16
Figure 10 Effect of variation on string length and module count per SMA 15kW inverter
The maximum string length for this Trina module to be within the 800V VMpp range of the inverter was 20 but
was not an attractive solution from a system output standpoint. The overload losses were 4.8% for this string
length and thus was not even worth considering. This iterative trial and error process to find the optimum string
length was done for each of the 4 module types. It can be seen that the optimum string length for the Trina
modules utilizing the 15kW SMA inverter is 17 modules per string which corresponds to the highest specific
output and performance ratio over the range of possible string lengths.
Table 5 shows the detailed inverter losses from the variation in string lengths in PVSYST. It must be noted that
in the case of 17 modules per string the inverter loss during operation decreased slightly due to reduced resistive
losses observed at higher operating voltages. The inverter loss over nominal power at 17 modules per string is
marginal at -0.1%. However it must be noted that at 19 modules per string this inverter loss over nominal power
jumps to -1.1%. This is as a result of the array being too large and the available array MPP energy exceeding the
inverter limited array energy. This can be seen in Figure 11 for the case of 20 modules per string. For 20
modules per string with an array oversizing of 1.67 a Voc at 18oC of 773V and a corresponding overload loss of
4.2% (energy which cannot be used by the inverter), this shows that even though the design limit is 1000V DC
and the inverter is rated for 1000V DC, this Voc limit cannot be approached due to the corresponding IV curve
being too far from the nominal inverter voltage of 600V. This type of array oversizing is not justifiable from an
output efficiency point of view however if modules continue to become substantially cheaper and space is NOT
a limitation then this level of array oversizing could be deemed justifiable to reduce BOS cost through the
reduction of the number of inverters required, however for this application it is not.
Table 5 Inverter losses for varying Trina 250W string lengths
Inverter Losses
75 modules / 85 modules
15kW inverter /15kW inverter
Inverter Loss during operation
-2.5%
-2.4%
(efficiency)
Inverter Loss over nominal inv Power
0.0%
-0.1%
Inverter Loss due to power threshold
0.0%
0.0%
Inverter loss over nominal inverter
0.0%
0.0%
voltage
Inverter Loss due to voltage threshold
0.0%
0.0%
17
95 modules /
15 kW inverter
-2.6%
-1.1%
0.0%
0.0%
0.0%
Figure 11 Array voltage & Power sizing for Trina 250W Maximum possible string length of 20 (Screen shot from PVSYST)
3.3.2 String Numbering
Once the string lengths were determined to satisfy voltage constraints of the inverter, the string numbers feeding
into the inverter and thus overall module count per inverter were determined. As mentioned in the previous
section , this inverter represents a unique design with 2MPP trackers, one for input A and one for input B. Input
A has flexibility of up to 5 strings with a maximum input current of 33A per input A. Input B only has flexibility
for one string and has a maximum input current of 11A. There is an electronic string fuse for Input A of 40A
and for Input B of 12.5A which cannot be exceeded. Taking into account these current constraints, the number
of strings utilized per inverter was determined for each module type based on the STC Impp value for each of the
module types. For example, Impp at STC for the Trina modules was found to be 8.01A. As such the number of
parallel strings which could be fed into Input A with a maximum input current of 33A was found as simply
33/8.01 which allows for 4 inputs to be used out of the available 5 on input A. Similarly for input B with a
maximum input current of 11A, only one string at 8.01A can be fed into this input. As such, due to these
current constraints, all channels for input A could not be utilized for all module types for example, Trina and
Ying Li modules only allow for the utilization of 4 channels per input A because of their higher Mpp currents.
This is shown in Table 6. The remaining channel on input A would be capped in the field. Table 6 shows the
current and voltage characteristics for the 15kW inverter utilizing Trina Modules. Since there are 4 strings of 17
for input A , this gives a total of 68 modules on input A and 1 string of 17 on input B thus giving a total number
of 85 modules per inverter. These inverter input details were done for Sanyo , Ying Li and First Solar modules
as well and can be found in Appendix E.
18
Table 6 Inverter Input Details for Trina-250 P05A module
Inverter Input Details for Trina-250 P05A
Sunny Tripower STP
Inverter Type
15000TL
Mpp voltage Range
360-800V
String Length
17
Strings Per MPP input A
4
Strings Per MPP input B
1
Total # of modules Per 15kW inverter
85
Inverter Details
Input A
Input B
Module Count
68
17
Max Inverter Vmpp
Max input Voltage, Voc at 18oC
Min String Voltage to be within Inverter Vmpp
Min Input Voltage Vmpp at 85oC
Maximum Input Current
String Input Current (Impp)
Maximum electronic Fuse Short Circuit Current
Inverter Short Circuit Current (Isc)
800V
657V
800V
657V
360V
419V
33A
32.04A
360V
419V
11A
8.01A
40A
34A
12.5A
8.5A
3.4 Preliminary Results
3.4.1 Inverter Level Summary
The next step was to run PVSYST simulations for the various module types per 15kW inverter, the lowest
system denominator. These simulations were done while keeping all parameters constant save module type and
of course module numbers per inverter. To get an idea of the relative performance between module types per
inverter the azimuth was fixed at -135◦ (North East Direction). The rationale behind this was that the majority
of modules in reality would be generally oriented in this direction. A tilt of 5◦ was fixed for the modules again
because the majority of modules to be installed in reality will be at this tilt. This tilt and orientation resulted in an
annual radiation on the plane of approximately 1927kWh/m2. The individual PVSYST simulations were run and
the data was then extracted and can be seen in Table 7. A sample of the individual PVYST reports can be found
in Appendix M.
19
Table 7 PVSYST simulation results for different module brands per 15kW inverter
Module
Modules per Inverter
Azimuth
Tilt(5 degrees)
Module Area(m2)
Nominal PV Power(kWp)
Maximum PV Power
Nominal AC Power(kW)
Array Oversizing
SANYO HIT
90
-135
5
125
21.6
20.3kWdc
15
1.44
TRINA
85
-135
5
139
21.25
19.0kWdc
15
1.42
YING LI
75
-135
5
146
21
18.5kWdc
15
1.4
FIRST SOLAR
220
-135
5
158
19.8
19.1kWdc
15
1.32
Specific
Production(kWh/kWp/yr)
1542
1494
1457
1598
Normalized Daily
Production(kWh/kWp/day)
Approximate Annual
System
Production(kWh/yr)
Performance Ratio
Subsystem Cost (USD)
4.22
4.09
3.99
4.38
33310
0.8
$29,160
31744
0.775
$17,000
30597
0.756
$14,700
31640
0.829
$18,810
Output per area
((kWh/yr))/m2)
$/kWh (20yrs)
266.48
$0.04
228.37
$0.03
209.57
$0.02
200.25
$0.03
It was expected for the SANYO HIT modules to have the highest specific yield due to them having the highest
efficiency (17.35%) under STC conditions as well as due to their reasonably high performance under elevated
temperatures. Surprisingly however as can be seen from Table 7, the First Solar CdTe had the highest specific
production of 1598 kWh/kWp/yr and highest performance ratio of 0.829 despite having the lowest module
efficiency of 12.54 %. First Solar’s CdTe superior specific production and performance ratio can be explained
due to two factors. Firstly relative temperature coefficients of Voc decrease with increasing band gap energy.
Typically, CdTe (1.44 eV) and amorphous Silicon (1.7 eV) have higher band gaps than crystalline modules (1.1
eV) and consequently have lower temperature coefficients. Secondly, modules with higher efficiency at STC
tend to have lower Rseries & high Rshunt which causes a reciprocal effect on efficiency at low irradiance levels thus
favoring CdTe & amorphous silicon at low irradiance levels. These array losses for the various module types are
depicted below in Table 8. When taking into account absolute performance only, the clear winner is First Solar’s
CdTe cell with the highest specific production and performance ratios and with the lowest total array losses of
12.6% as can be seen in Table 8.
Table 8 Array losses for different module brands per 15kW inverter
Array Losses
PV loss due to
irradiance level
PV loss due to
temperature
Module Quality
Loss
Module Array
Mismatch Loss
Ohmic Wiring Loss
Total Array losses
SANYO HIT
TRINA
YING LI
FIRST SOLAR
-1.5%
-3.1%
-3.0%
-0.7%
-8.7%
-12.4%
-13.9%
-7.4%
-2.6%
-0.1%
-1.6%
-2.6%
-2.1%
-2.1%
-2.1%
-1.0%
-0.9%
-15.8%
-1.0%
-18.7%
-1.0%
-21.6%
-0.9%
-12.6%
20
When available area is factored in however which is generally a constraint on a roof top, we see from Table 7
that SANYO HIT performs the best in terms of output per unit area. Taking this one step further and including
cost as a constraint, we see that Ying Li has the lowest module cost as well as the lowest cost per kWh (over
20years). Based on these different results and depending on the perspective taken, it is very hard to choose the
“best” module at this level of analysis. As such, these results were scaled up to the system level to look at cost
and performance considering the BOS costs. In addition RESCO’s financial model was used to determine what
the bests system would be from a NPV, Payback period and ROI perspective which the client would be mostly
concerned with from an investment standpoint.
3.4.2 System Level Summary
Table 9 represents the system level summary of the different Systems along with the corresponding percentage
load offsets for each system. This global system size for each case is based on the maximization of the available
roof area as well as the total number of modules for each system being divisible by the number of modules per
inverter in order to give an inverter count which is an integer.
Table 9 System summary of various PV Systems based on module brands
Module
Sanyo
HIT
Trina
Ying Li
First
Solar
Total #
of
modules
Module
Area
(m2)
# of
15kW
inverters
Nominal
PV
Power
(kWp)
Projected
Annual
System
output
(kWh/yr)
Total annual
load(2012)
(kWh/yr)
% of
Carlton's
load
offset
2160
1870
1575
3000
3058
3066
24
22
21
518.4
467.5
441
799,372.80
698,445.00
642,537.00
2,259,130
2,259,130
2,259,130
35.38
30.92
28.44
4180
3002
19
376.2
601,167.60
2,259,130
26.61
The next task was to determine the total system cost and price per watt for each of the various systems. The
exact system sizes didn’t matter as much in this analysis as performance metrics taken from PVSYT were per
Wp. In addition, most BOS systems cost were given per Wp as well. Several sources were used for pricing of
components including sales representatives, manufacturer and supplier’s price sheets, Resco’s Metrics as well as
online quotation tools for example in the case of racking. The cost tables include shipping costs embedded in
the cost of each component. The cost table comparison is shown below in Table 8. It must be noted that a
higher prices per watt for racking was used for thin film in addition to the fact that there are significantly more
modules in the case for the CdTe and therefore greater quantities of mid clamps, L-feet and other mounting
components will be required thus driving up the overall mounting cost. Similarly, due to the high module count
of first solar (4180 modules), the cost of labor per watt was increased due to the modules decreased wattage per
module compared to the other technologies. Based purely on system cost we see that Ying Li is the best choice
from an initial investment perspective. However in the following section, this relative price is compared with the
performance of the system to determine the best overall choice.
21
Table 10 Cost table for various PV Systems based on module brand
Sanyo HIT Cost Table
Price per W p (USD)
Module
$1.35
Inverter
$0.28
Racking
$0.35
AC +DC wiring
$0.15
Inveter 20yr Warranty
$0.07
Miscellaneous Components
$0.05
Equipment Rentals
Inverter Shade Structure
Webbox System monitoring
Labour
$0.25
Total Cost
Engineering
5% of Total Cost
Profit
20% of Total Cost
VAT
17.50%
Sales Price
$3.52
Total Cost (USD)
$699,840.00
$101,833.92
$181,440.00
$77,760.00
$33,936.00
$25,920.00
$10,000.00
$12,000.00
$1,500.00
$129,600.00
$1,273,829.92
$63,691.50
$254,765.98
$234,066.25
$1,826,353.65
Trina Cost Table
Price per W p (USD)
Module
$0.80
Inverter
$0.28
Racking
$0.35
AC +DC wiring
$0.15
Inveter 20yr Warranty
$0.07
Miscellaneous Components
$0.05
Equipment Rentals
Inverter Shade Structure
Webbox System monitoring
Labour
$0.25
Total Cost
Engineering
5% of Total Cost
Profit
20% of Total Cost
VAT
17.50%
Sales Price
$2.74
Total Cost (USD)
$374,000.00
$93,347.76
$163,625.00
$70,125.00
$31,108.00
$23,375.00
$10,000.00
$11,000.00
$1,500.00
$116,875.00
$894,955.76
$44,747.79
$178,991.15
$164,448.12
$1,283,142.82
Ying Li Cost Table
Price per W p (USD)
Module
$0.70
Inverter
$0.28
Racking
$0.35
AC +DC wiring
$0.15
Inveter 20yr Warranty
$0.07
Miscellaneous Components
$0.05
Equipment Rentals
Inverter Shade Structure
Webbox System monitoring
Labour
$0.25
Total Cost
Engineering
5% of Total Cost
Profit
20% of Total Cost
VAT
17.50%
Sales Price
$2.61
Total Cost (USD)
$308,700.00
$89,104.68
$154,350.00
$66,150.00
$29,694.00
$22,050.00
$10,000.00
$10,500.00
$1,500.00
$110,250.00
$802,298.68
$40,114.93
$160,459.74
$147,422.38
$1,150,295.73
First Solar Cost Table
Price per W p (USD)
Module
$0.95
Inverter
$0.28
Racking
$0.40
AC +DC wiring
$0.15
Inveter 20yr Warranty
$0.07
Miscellaneous Components
$0.05
Equipment Rentals
Inverter Shade Structure
Webbox System monitoring
Labour
$0.30
Total Cost
Engineering
5% of Total Cost
Profit
20% of Total Cost
VAT
17.50%
Sales Price
$3.14
Total Cost (USD)
$357,390.00
$80,618.52
$150,480.00
$56,430.00
$26,866.00
$18,810.00
$10,000.00
$9,500.00
$1,500.00
$112,860.00
$824,454.52
$41,222.73
$164,890.90
$151,493.52
$1,182,061.67
3.5 Preliminary Financial Analysis
Resco has a financial tool which is instrumental in estimating the financial viability of its projects from the initial
stages via certain financial metrics such as ROI and NPV. This tool was adapted to differences in market
conditions and regulatory and licensing framework for conditions in Barbados. There are significant
assumptions which were used in determining the financial success of these different system types and these
assumptions are listed below:
• The Project life time is 20 years.
• The system will be metered at a cost of electricity of $0.35/kWh
• Annual Reporting and Monitoring Cost are 0.5% of Annual Savings
• Annual Operating & Maintenance cost are 4.0% of Annual Savings.
• Annual Insurance Cost is 0.25% of Total Capital Cost.
• Annual production degradation in output is 0.75%.
• A discount rate of 10% is assumed.
• There a no inverter replacement cost as the cost for 20 year warranty was included in system cost as
shown above in Table 8.
• The project funding is 100% equity financed.
22
Table 11 Financial analysis for various system types
SANYO
Solar Project Financial Analysis
Total Equity Investment
Net Present Value
Simple Payback
Estimated Rate of Return
Avg. Yearly Free Cashflow
Cumulative 20yr kWh generated*
LCOE( $/kWh)
YING LI
Solar Project Financial Analysis
Total Equity Investment
Net Present Value
Simple Payback
Estimated Rate of Return
Avg. Yearly Free Cashflow
Cumulative 20yr kWh generated*
LCOE( $/kWh)
$1,826,353.65
$299,030.00
7.1 Years
12%
$243,563.00
14,898,014
$0.265
TRINA
Solar Project Financial Analysis
Total Equity Investment
Net Present Value
Simple Payback
Estimated Rate of Return
Avg. Yearly Free Cashflow
Cumulative 20yr kWh generated*
LCOE( $/kWh)
$1,283,142.82
$579,127.00
5.7 Years
16%
$213,589.00
13,017,010
$0.210
$1,150,295.73
$562,822.00
5.5 Years
17%
$196,566.00
11,975,045
$0.204
FIRST SOLAR
Solar Project Financial Analysis
Total Equity Investment
Net Present Value
Simple Payback
Estimated Rate of Return
Avg. Yearly Free Cashflow
Cumulative 20yr kWh generated*
LCOE( $/kWh)
$1,182,062
$419,998.42
6.1 Years
15%
$183,649.11
11,204,038
$0.226
* assumes an annual degradation in
output of 0.75%/yr for 20 yrs &
discount rate of 10%
After taking total system cost into consideration as well as performance, it can be seen that at 17% ROI, Ying Li
has a slightly better rate of return than Trina and First Solar modules. Sanyo is way out of the picture at
$1.35/Wp. Trina, nd Ying Li all have similar payback periods of less than 6 years. Trina has the highest Net
Present Value out of the four options. Ying Li and Trina both have very attractive levelized cost. However,
when taking when taking into consideration the % of Carlton’s Load offset as well as performance and cost it is
believed that the Trina PV system has the best all-round performance with a higher PR than Ying Li as well as a
greater percentage of load offset. As such, the Trina Modules were taken through the detailed design process.
One additional key note is that all 4 systems LCOE is still well below the cost of electricity to the user currently
paid at $0.35/kWh
3.6 Racking Selection
The racking design constraints were not as rigorous for the racking system as compared with the module and
inverter selection. The main requirement for the racking was for it to be able to withstand 130mph winds as
outlined in the boundary conditions and therefore required a certain thickness and robustness. After careful
consideration of a variety of brands it was decided to go with DPW P6 Power Rail racking due to their cost
competitiveness and the fact that there is a supplier based in Puerto Rico making shipping and logistics easier.
The design specification is included in Appendix K. The main consideration from the specification sheet was
the Rail Span and maximum cantilever. These were determined to be 4 feet and 2 feet respectively. Figure 12
shows the detail of the trapezoidal metal roof of Carlton and Figure 13 shows how the intended installation of
the racking will look. There will be a roof block which sits on the ridge through which L-foot penetrations will
be made through to the purlins below. The L-foot will then be bolted down and sealed with washers and
polyurethane caulking to ensure each penetration is leak tight. The roof block is a necessary component for this
roof type since it allows for roof attachment without crushing or collapsing the ridges of the roof material. A
detailed cross-section of this assembly is shown in Appendix K. The Trina 250 P05A installation manual
provided details on the spacing requirement between rails provided that these modules will be installed in
portrait. This distance between rails was determined to be 825 mm and has been captured in the racking layout
drawing which is in the Results section. The Trina Installation manual can be found in Appendix L.
23
Figure 12 Picture of Carlton Roof Section D detailing Roof Material
Figure 13 Picture showing proposed mounting for Carlton’s trapezoidal roof (Snapnrack, 2013)
4 Detailed Design
In order to have a constructible project which can be successfully built with minimal technical
challenges in the field, it is necessary to formalize a stepwise plan which flows from the initial
conceptual design through to construction while accounting for every intermediary step along
the way. The following steps which are deemed necessary for detailed design are explained below
in terms of what sub tasks are generally encompassed within these steps and why they are
necessary in the design stage. Generally, the engineering design approach taken should account
for the following steps:
•
Feasibility & Site Visit
24
•
•
•
•
•
•
•
Submission & Approval of Interconnection Agreement & review of Utility
Requirements & NEC Code
Structural Evaluation (outside the scope of this thesis)
Shading Analysis
Detailed Roof Layout
DC Design
AC Design
BOS Take offs (done by Resco)
The list of deliverables or engineering drawings which are included in the design scope of this
thesis and which shall be issued prior to construction are as follows:
•
•
•
•
•
•
•
Site plan & Site Coordination Plan
Structural Report on roof loading (will be issued but is outside the scope of this thesis)
Array Layout
Module Interconnection Drawing (MID)
Single Line Diagram (SLD)
Conductor Equipment Schedule
Bill of Materials (BOM)
4.1 Feasibility and Site Visit
This first visit is crucial for commercial applications. It is intended to get as much detail as
possible on the proposed location before design commences. This is a logistic site visit, where
major roof obstructions and shading objects are noted which would be taken into account during
the design of the array layout. Other things which are taken into consideration are location of the
electrical room and proximity to the proposed inverter so that wire lengths which can be costly
are minimized. The electrical room must be examined for metering and disconnect logistics and
placement. Panoramic photos are also taken around the site which will help determine site
logistics and where various activities will take place on site for example where craning will likely
take place, where the stair tower will be placed, where the site trailer will be located etc. Any
logistical problems can be identified at any early stage which can help avoid costly mistakes
during construction. Such issues which would have to be considered are; proximity of the
building to high voltage lines which would affect where craning of materials is likely to occur.
Also truck loading bays and their proximity to roof access points.
A site visit was conducted in January of 2013 and the above factors were considered and noted
for the Carlton site. A number of pictures were collected of the location and of the roof top
which assisted in the creation of the Site Coordination plan shown in Appendix Q.
4.2 Submission of Interconnection Agreement
This process is unique for every country and is utility dependent. In Barbados this process
involves the completion of an interconnection application. This form defines what type & size of
renewable energy generator will be connected to the grid in addition to the system details
(equipment specifications) as well as the type of grid connection proposed. A completed
interconnection form can be found in Appendix B. There are two configurations permitted by
25
the utility; 1) Net generated electricity being fed to the grid 2) Total electricity production fed to
the grid. These arrangements are depicted below in Figure 14 & Figure15 respectively. The
current utility restriction for feeding into the grid is 150kWp. Due to the proposed PV system
size for Carlton of 425kWp, this restricts the connection type to “net generated electricity to the
grid”, whereby the PV system will be fed into the building’s Main Electrical Panel.
Providing that this approval is granted from BL&P to connect, the client then fills out an
interconnection agreement form which outlines the requirements to which the customergenerator must adhere in order to be compliant with the utility’s processes and power quality
requirements. Also needed is a Power Purchase Agreement form which is a contract outlining an
agreement on the amount and duration of the Feed In Tariff (FIT) which the utility is committed
to paying the customer-generator providing they are approved for grid interconnection.
Figure 14 Net Electricity Production to the Grid (BL&P, 2011)
Figure 15 Total electricity production to the grid (BL&P, 2011)
26
4.3 Structural Evaluation
This report is necessary for commercial systems as a sizeable PV system will add substantial load
to the existing roof structure. This is one of the fundamental requirements (structural capacity)
from a safety standpoint. This step ultimately determines the viability of the project and can
affect the type of modules & racking used as well as indicate structural exclusion zones where
modules can’t be placed due to suspended roof loading, or lack of reserve roof capacity. This
step is especially crucial for ballasted roofs which were designed to have a certain load to
maintain roof integrity. This step is outside the scope of this thesis; however it will be completed
if this project gets the go-ahead.
4.4 Shading Analysis
An important step in the optimized designing of any PV System whether residential or
commercial is performing a shading analysis on the proposed site. It is well known that shading
of a module can significantly reduce the output of that whole string it is on because of the
decrease in voltage and current. The effect of shading depends on how much of the module is
shaded, where the module is shaded (which of the by-pass diodes is activated) and whether the
inverter has Maximum Power Point Tracking to reduce the inverter output losses.
Figure 16 shows shading which occurs on Section F of Carlton’s roof at mid-morning due to the
wall surrounding the AC equipment. Similarly, an area of Section E is shaded in the mid
afternoon as well as Section G, but to a lesser extent. Because of this known shading potential
for loss in output it was important to do a detailed shading analysis in PVSYST for Carlton to
determine how much of an offset from the wall was required to avoid shading and to determine
how to string modules which were shaded outside of peak sun hours.
Figure 16 Shading of Roof Section F of Carlton during mid-morning
27
The Carlton model was constructed in PVSYST by using accurate dimensions from existing site
plans and elevations. The created model is depicted in Figure 17. The stringing of the modules
to the inverters (MID) was designed based on the shading tool in PVSYST for the Winter
solstice of Dec 21st 2013. The layout was based on putting shaded modules on input B of the
various inverters which has its own MPP tracker so that the other 4 strings per inverter would
not be affected from the one shaded string. The shading model constructed and used in
PVSYST is shown below. Simulations were done for summer and winter solstices as well as
spring and fall equinoxes. Total beam losses incurred over a day at the Spring Equinox (March
20th 2013) was 1.7%. This loss was 1.2% at the summer solstice (June 21 2013) and 1.7% and the
fall equinox (September 22nd 2013). The greatest beam losses due to shading occur at winter
solstice (Dec 21st 2013) as expected and amount to 2.9%. This simulation is shown in Figure 18.
Even though shading losses only amount to 2.9% during this time, these losses are minimized by
optimizing the array layout and the string layout accordingly.
Figure 17 PVSYST Shading Model for Carlton A-1
28
Figure 18 Carlton A-1 shading loss after day of Simulation of Dec 21st 2013
4.5 DC Design
The DC design involves the selection of module type, layout, orientation, array size and string
lengths & numbers as well as combiner box sizing (if required). In essence this design
encompasses all aspects of design up until the inverter. The DC design also determines DC wire
lengths required from combiners to the inverter and then determines the corresponding size
based on required ampacity of the conductor & the size based on allowable voltage drop from
the combiner box to the inverter.
The scope of the DC design for this thesis included determining string length, system size,
module spacing and orientation, the Module Interconnection Diagram, DC voltage drop from
the homerun to the inverter, racking selection and design.
4.6 AC Design
The AC design consists of the Single Line Diagram identifying wire sizes throughout the system
from the DC side through to the combiner, recombiner, inverter and to the meter & main panel
board. These sizes are determined by calculating the ampacity of the wire. It also consists of the
conductor schedule which illustrates wire characteristics such as rated current, NEC de-rated
current, conductor size, over current protection, bonding conductor size as well as number of
conductors. These calculations take into consideration the maximum allowable system voltage
drop and as mentioned previously, the derated current and overcurrent protection as specified by
the Governing Electric Body.
29
The scope of the AC design for this thesis includes, the sizing and placement of inverters on the
roof, the voltage drop calculations from the inverter to the AC combiner panel, voltage drops
from the AC combiner panel to the building main electrical panel, conductor schedule, sizing of
all breakers fuses and AC disconnects according to NEC code and utility requirements and
creation of a SLD.
4.7 BOS Takeoffs
Generally BOS refers to all other costs outside of the module cost and therefore includes
mounting structure and associated cost, labor, engineering work, project management. This can
also include land if the space is rented. BOS costs can be further broken down into BOS
Equipment Cost and BOS Soft Cost, where Soft cost refer to profit, installation labor,
permitting, inspection and interconnection fees, arranging third party financing & customer
acquisition(Adrani et al. 2012). In particular, customer acquisition cost which are related to bid
and proposal generation, contract negation and design are expected to represent relatively high
sunk costs to the installer in Barbados and other Caribbean Islands where the PV market is more
infantile than in Europe or the North America where the industry is more established.
Figure 19 shown below provides the installer with a good snapshot of the relative breakdown of
BOS system cost per watt for a utility scale project in the USA which serves as a good marker to
whether BOS projected cost are reasonable and where these cost can likely be optimized.
Miscellaneous costs include soft BOS costs, engineering, project management etc. Although the
data for this figure is taken from utility scale projects which would have relatively lower BOS
system cost/W than commercial due to economies of scale effect this figure still provides a good
rough target for BOS system cost likely to be incurred for Carlton.
Figure 19 BOS Cost Breakdown 10MW, Fixed Tilt Multi c-Si Project in the US (Aboudi, 2011)
For the final system size, the BOS estimates were done by Resco’s estimating team based on the
created MID, array layout and conductor schedule which are shown in the engineering drawings
section. This BOS breakdown was crucial in determining the final project cost to the client.
Detailed breakdown of the BOS costs can be found in Appendix F.
30
4.8 Final System Simulation
The final system design was comprised of 1700 Trina PA05 250W modules utilized across 20
inverters at 17 modules per string and 5 strings or 85 modules per inverter. The pdf version of
the site plan was converted and imported into AutoCAD. AutoCAD was used in coming up with
a final module count. The module dimensions were modeled in AUTOCAD and strategically
placed across the roof to maximize the available roof space. The correct building azimuth was
also determined at this point to be -126o. This was done using via geo-referencing the location in
to AutoCAD. Preliminary estimates based on existing drawings were that the building was at 135o.The geo-referencing of the building site acts for more accurate PVSYST simulations. This
final number of 1700 modules resulted from taking into account appropriate offsets from the
building edges, generally one meter in most cases. These offsets were also maintained between
roof sections to allow for suitable walkways thus facilitating relatively easy roof section access by
personnel if required. This module count also results from a module spacing of 3/8” using the
DPW mid clamps. In addition, greater offsets were used in placement of modules around the
base of the wall enclosing the AC equipment in section E. F and G. Larger offsets were made for
section E and F where the front and back roof sections meet as in order to avoid shading of
sections E and F caused by the southern roof sections.
A total of 11 simulations were run in PVSYST, some of which were modeled utilizing the
heterogeneous orientation allowing for two different tilts and orientations to be modeled using
the same inverter. These results are shown below in Table 12.
Table 12 shows an overall weighted system performance ratio of 76.3 % and a weighted specific
production of 1467kWh/kWp which represent respectable performance metrics after taking into
account shading losses and the fact that modules are utilized on roof sections which are not
ideally positioned for maximum irradiation. A total annual yield of 623,627kWh/yr allows for an
offset of approximately 28% of Carlton’s load based on 2012 electricity usage. Based on a cost of
electricity of $0.35/kWh, this annual production results in a gross saving of approximately
$218,269.00 USD per year. However, this savings doesn’t account for the operating and
maintenance cost as well as monitoring cost associated with the system. A more in-depth
financial summary is given below in the financial section.
The final simulation in Table 12 was the most important step of design as it projects how the
final 425kWp system designed for Carlton will perform. Table 12 is especially important as it
shows the breakdown of simulations run in PVSYST to accurately reflect the proposed design. A
total of 11 simulations were run and the results of the projected annual outputs added to give a
final annual system output. The final system performance ratio and specific output were
determined by taking the weighted average of the 11 individual simulations. The table reflects the
wiring of the MID and shows how the modules from various roof sections are connected across
inputs A and B for each of the 20 inverters which constitute the system. Table 12 also
demonstrates the sharing of an inverter by modules from different roof sections which was a
difficult task as much consideration had to be given to the proximity of modules to the inverter,
whether or not the modules would be shaded and that the total number of modules per roof
section were divisible by the module string length of 17. The MID explains this in greater detail
in the results section. A sample of the simulation results is given in Appendix N.
31
4.9 FINAL SIMULATION RESULTS
Table 12 Final System PVSYST Simulations
*indicates weighted average
Module Description
Tilt(o)
Azimuth (Q
Q)
Specific
Simulation
# of
#
inverters
1
3
2
1
3
1
4
2
5
1
6
1
7
2
8
2
9
4
10
2
11
1
Input A
204 from Section A across3
inverters
68 from Section B across 1
inverter
68 from Section C across 1
inverter
136 from Section D across 2
inverters
68 from Section D across 1
inverter
69 from Section D across 1
inverter
136 from Section E across 2
inverters
136 from Section F across 2
inverters
272 from Section G across 4
inverters
136 from Section G across 2
inverters
68 from Section G across 1
inverter
20
1360 modules globally
across 20 inverters
Total
Input B
Input A Input B
51 from Section A across 3
inverters
54
54
17 from Section A across 1
inverter
-36
54
17 from Section A across 1
inverter
144
54
34 from Section D across 2
inverters
-126
-126
17 from Section A across 1
inverter
-126
54
17 from Section F across 1
inverter
-126
144
34 from Section E across 2
inverters
-36
-36
34 from Section F across 2
inverters
144
144
68 from Section G across 4
inverters
-126
-126
34 from Section E across 2
inverters
-126
-36
17 from Section F across1
inverter
-126
144
Input A Input B
Nominal AC
Module Count DCKWp Power (kW)
Production(kWh/k Estimated Annual Performance
Wp)
Yield (kWh/yr)
Ratio
15
15
255
63.75
45
1506
96,000
0.773
15
15
85
21.25
15
1509
32,060
0.77
15
15
85
21.25
15
1438
30,560
0.771
15
15
170
42.5
30
1440
61,200
0.773
15
15
85
21.25
15
1460
31,020
0.777
15
5
85
21.25
15
1373
29,180
0.732
5
5
170
42.5
30
1468
62,400
0.747
5
5
170
42.5
30
1465
62,300
0.761
5
5
340
85
60
1478
125,647
0.765
5
5
170
42.5
30
1455
61,800
0.751
5
5
85
21.25
15
1480
31,460
0.767
1700
425
300
1467.3*
623,627
0.763*
340 modules globally
across 20 inverters
32
5 Results
5.1 Engineering Drawings
Below are some of the engineering design of Carlton’s PV System. The larger detailed
engineering drawings are listed in Appendix O-T along with additional engineering drawings for
construction which should be referred to for the comprehensive understanding of this section.
5.1.1 Array Layout
All twenty inverters were strategically placed in order to minimize the distance homerun strings
had to be wired which consequently results in lower wiring cost and reduced DC voltage drops
from the modules to the inverters. Cable lengths from the inverters to the electrical room were
calculated using Autocad for each of the 20inverters. The cable runs include a 10m vertical run
from the roof into the electrical room where the AC panel Board will be housed. These cable
runs are depicted in The Array Layout Drawing included in the engineering drawings section as
well as Appendix P. The inverters will all be wall mounted with a semi shade structure over them
to avoid inverter derating due to over temperature which can be experienced during midmorning to early afternoons in Barbados on a clear day. The conductor sizes were then
determined for the AC runs for each inverter using the voltage drop calculator from Southwire
(Southwire Company, 2005). These voltage drop calculations were done with a constraint of
having a maximum allowable voltage drop from each inverter to the electrical room of 1.55%. A
screen shot is shown below for the conductor sizing of inverter #4 from the Conductor
schedule in Appendix T.
Figure 20 Voltage drop calculation for Inverter #4 to AC Panel Board (Southwire Company, 2005)
33
34
5.1.2 Single Line Drawing
The SLD shown in Appendix S was developed by integrating the inverter specifications and designed Trina
PA05 module layout into an appropriately rated AC Panel Board. From the Panel Board the two conductors go
into a disconnect which prevent any harmful reverse currents from damaging the inverters. The AC Disconnect
DG2 will be located in accessible area for access of BL&P personnel in the event of grid maintenance which
would require the Solar Generator production to the grid to be interrupted. From DG2, the conductors are then
fed into a generation meter which will be located in the electrical room. This then feeds into disconnect DG1
which will also be located in the electrical room. The purpose of having two disconnects on either side of the
meter is so that the meter can be isolated and serviced or replaced if need be without risk of electrical shock to
personnel. From DG1 current is fed into the main panel through a 600A circuit breaker. This electrical
configuration of feeding into the main panel is such that electricity flows from the generator through the main
panel to the load and excess energy generated outside of meeting Carlton’s load is fed to the grid. This is in
compliance with BL&P’s allowable configuration of “net energy to the grid” (The Barbados Light & Power
Company Limited, 2011). If all electricity generated was to be fed to the grid, an electrical connection would go
from the RGS production meter to the line side of the building main breaker.
It must be noted that a crucial design requirement for the sizing of the conductors is that all wires are rated to
carry the current of the OCPD. This is so that in an overcurrent event, the OCPD fails before the conductor
which could cause an electrical fire or electrical shock to personnel. The circuit breakers are all sized for 1.25
times rated current as per NEC code requirements (National Electric Code, 2010). The AC Panel Board acts to
combine the AC output currents from the inverters into two conductors which feed into the fused AC
disconnect DG2. The AC Panel Board output breaker is rated to handle a combined current of 1.25 times the
20 inverter outputs each with a rated current of 24A. The details of the AC Panel Board are shown in the SLD.
As per SMA inverter specifications, input A is electronically fused at 40A while input B is fused at 12.5A, this is
depicted in the SLD in Appendix S.
5.1.3 Conductor Schedule
The conductor schedule was designed to account for cable lengths, individual conductor rated currents and
voltages along with NEC derated current requirements. The overcurrent protection is based on the next
available protection size up from the NEC derated current. The conductor is RW90, and conductor size is
determined by voltage drop and cable lengths using Southwire (Southwire Company, 2005). The number of
conductors is dependent on the current & voltage output as well as the phase; this is 3 + GRND for the
conductors leaving the inverter, 6 + GRND for the conductors leaving the AC Panel Board to the Meter. From
the Meter to the Main Panel there are 8 conductors + GRND (which is two parallel 4 wire runs). The extra
conductor times two is for the neutral which is run from the Meter and makes it easier to measure the voltage
across the meter and. The 4 wire is also consistent with the wiring to the main Panel from the Utility.
Conductors are run in parallel from the AC panel as indicated in the conductor schedule because of the higher
flowing currents after combining all twenty inverter outputs and thus the required higher ampacity which could
not be carried in one single 3 + GRND conductor. Therefore the current is split into two parallel runs. The PV
string homeruns have 2 + GRND conductors. The bonding conductors depicted in the conductor schedule of
Appendix T are all sized according to bonding requirements stipulated in the NEC Code (National Electric
Code, 2010). TECK is used as the raceway material outdoors which is armored and well insulated suitable for
outdoor conditions. EMT which is metal guarded and is not as robust as TECK will be used indoors.
35
36
5.1.4 Module Interconnection Drawing
The MID identifies how each string is connected within the array. It determines the positive and negative
terminals of each string and how these are all run to the inverter. This task proved to be one of the more
challenging tasks in completing as several factors had to be considered simultaneously when stringing together
the modules. The factors involved in the determination of the string layout are as follows: 1) Proximity of the
modules to the inverter to minimize DC voltage drops. The 10AWG wire used for homeruns did allow for some
string design flexibility allowing up to 150 feet (45 m) for a voltage drop of less than 0.5 %. Nevertheless the
strings were started and ended so that the terminals (+,-) were as close as possible to the inverters thereby
minimizing the length of cable required which ultimately helps to keep BOS cost to a minimum.2) Another
critical factor was that of shading and the desired effect of stringing the modules in a way that this shading
affected the overall system output as little as possible. To this end, Sections E & F which are the most shaded
sections had as many of the shaded modules place on Input B of the inverters with their own MPP tracker thus
effectively minimizing output losses. Lastly, although each roof section’s module count was divisible by 17, it
didn’t necessarily mean that these sections were divisible by a full inverter ie 85 modules (17 across 5 inputs; 4
for input A, 1 for input B.) As such, careful thought went into what string had to be borrowed from an adjacent
section for input B. An example of this type of stringing is depicted in the MID for inverter #4 in which 68
modules come from Section B, but the remaining 17 which are stringed on Input B come from Section A in
order to make up a full inverter. By determining which modules are shaded, which sections require extra strings
to fully utilize an inverter and stringing modules within close proximity of inverters, the resultant effect is an
optimized string design as was achieved and is depicted in the MID drawing of the engineering drawings and
Appendix R which is broken up into Sheet 1 and Sheet 2 for better visibility.
For nomenclature of the strings, the prefix represents the inverter number to which the module belongs, letter A
or B determines which input the module resides on and the digit 1-4 determines which string on input A the
module belongs to. For example 5-A3 is indicative of a module lying on string 3 of input A on inverter number
5. It must be noted that each string consists of 17 modules throughout the entire design.
37
38
5.2 Project Costs
With the help of Resco’s estimating and costing spreadsheet, one was adapted for Carlton based on its
uniqueness. Total Project Cost were comprised of 4 main costs; 1) Material Cost in the form of a BOM, 2)
Labor Costs which are comprised of direct and incidental labor, 3) General Expenses and 4) Engineering Costs.
The cost breakdown can be found in Appendix F. On top of these costs, shipping is included at a conservative
estimate of 5% of the total material cost. Mark up on the project is calculated at 20% of the total cost and VAT
is also included in the final project cost at a rate of 17.5% of the selling price. Prices throughout the various
tables shown below are given in CAD until the end where the final project cost of $1,099,599 is given in USD.
An exchange rate of 1USD =1.032 CAD was used for the final project cost. This equates to a cost per watt for
the 425kWp system of $2.59/W which is used in determining the financial metrics.
General Expenses cover cost which include items outside of direct material cost and include such items as Sea
Cans, Stair Tower, Portable Toilets, Permitting, Lamacoids etc. These costs are captured in
Table .
Engineering Cost Include full system design, structural analysis and all third party work subcontracted out.
These costs encompass the engineering disciplines of mechanical, electrical and structural.
Table 13 Final pricing of the proposed 425KWp PV System for Carlton A-1
FINAL PRICING
Final Pricing
Database Material (Extension)
Material Total
Shipping Cost @ 5 % material Cost
Direct Labor
Incidental Labor
Labor Total
Subcontractors
General Expenses
Total Cost
Engineering Cost @ 6% total cost
Total Markup @ 20% total cost
Selling Price
VAT @ 17.5%
Final Price CAD
Final Price in USD
Value ($)
614,553.55
$614,553.55
30,727.68
74,510.52
30,987.72
$105,498.24
$15,688.00
766,467.47
$45,988.05
153,293.49
$965,749.01
169,006.08
$1,134,755.09
$1,099,599.24
5.3 Financial Model
Using Resco’s financial model with all the same assumptions as those outlined in the preliminary financial
analysis, the project provides a favorable ROI of 17%, a simple payback period of 5.4 years and a NPV of
$563,731. The discounted payback period with a discount rate of 10% is found to be an attractive 8.3 years.
These metrics are depicted below in Table 14.
The LOCE was found to be $0.201/kWh and this figure which was estimated conservatively, was determined by
the sum of the annualized capital cost plus the annualized operating and maintenance cost plus the annualized
cost for monitoring as well as the annualized cost of insurance. This sum was then divided by the average annual
39
production of the system over 20years of 581,385 kWh/yr which accounts for a system derate in output of
0.75% annually.
Table 14 Financial Analysis Summary for 425kWp PV System proposed for Carlton A-1
TRINA
Carlton A-1 Financial
Analysis Summary
Total Equity Investment
Net Present Value
Simple Payback
Discounted Payback
Estimated Rate of Return
Avg. Yearly Free Cash flow
Cumulative 20yr kWh generated*
LCOE/kWh
The Project life time is 20 years.
The system will be metered at a cost of electricity of $0.35/kWh
Annual Reporting and Monitoring Cost are 0.5% of Annual Savings
Annual Operating & Maintenance cost are 4.0% of Annual Savings.
Annual Insurance Cost is 0.25% of Total Capital Cost.
Annual production degradation in output is 0.75%.
A discount rate of 10% is assumed.
There a no inverter replacement cost as the cost for a 20 year warranty was included in system cost.
The project funding is 100% equity financed
$3,000,000.00
$2,500,000.00
$2,000,000.00
$1,500,000.00
USD
•
•
•
•
•
•
•
•
•
$1,099,599
$563,731
5.4 Years
8.3 Years
17%
$190,907
11,627,705
$0.201
Cumulative
Cash Flow
$1,000,000.00
$500,000.00
$0.00
-$500,000.00
-$1,000,000.00
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Fiscal Year
-$1,500,000.00
Figure 21 Cumulative Cash Flow of a 425kWp system over 20 years at a fixed cost of electricity of ($0.35/kWh)
40
$4,500,000.00
$4,000,000.00
$3,500,000.00
$3,000,000.00
$2,500,000.00
Cumulative Cash Flow
with annual electricty
escalation in cost of 3%
$2,000,000.00
$1,500,000.00
$1,000,000.00
$500,000.00
USD
$-
$(500,000.00)
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
$(1,000,000.00)
Fiscal Year
$(1,500,000.00)
Figure 22 Cumulative Cash Flow of a 425kWp system over 20 years at a elecricity escalation rate of 3% annually
Figure 21 and Figure 22 show the projected cash flows over the life time of the project for a fixed cost of
electricity, and an annually 3% escalated cost of electricity respectively. Figure 21 represents a very conservative
approach as the cost of electricity in Barbados has more than doubled from 2009 to 2012 (Barbados Economic
& Social Report, 2012). This table showing this drastic increase in electricity to the consumer can be found in
Appendix G. Even with this overly conservative estimate (assuming the cost of electricity is fixed at $0.35/kWh
for 20 yrs) the client will have accumulated USD $2,718,531.00 worth of savings over the project lifetime. When
an annual rate increase of 3% is attached to the cost of electricity, at the end of 20 years the client will have
accumulated USD $4,076,401 or BDS $8,152,802 which equates to an average annual savings of BDS
$407,640.00 over the course of 20 years. Adding this 3% rate increase in electricity also increases the NPV to
$953,952 USD, and the ROI to 20% over the 17% calculated at a fix electricity rate of $0.35/kWh.
5.4 Proposed Incentives
A full list of proposed Government incentives for Renewable Energy generators and investors in renewable
energy can be found in Appendix H. These incentives were proposed in Parliament in June of 2012(Nation
News Barbados, 2012). More recent developments to have these incentives implemented have been carefully
reviewed by Government and negotiated with the Central Bank through Finance and Economic Affairs
Minister, Christopher Sinckler of Barbados with the aim of having these incentives legislated. This proposed bill
is likely to be passed after the August 13 2013, delivery of the Financial Statement and Budgetary Proposals (The
Barbados Advocate, 2013). Included in the forthcoming legislation will be a 10 year tax holiday for renewable
energy installers and developers, eligible businesses will be given up to 150% deduction on the amount of
interest paid on a loan relating to renewable energy (The Barbados Advocate, 2013). 150% deduction on taxable
income earned for those businesses involved in the development of renewable energy (The Barbados Advocate,
2013). Additionally there will be 150% taxable income deduction on the marketing of renewable energy (The
Barbados Advocate, 2013).
6 Discussion
One of the main goals at the outset of this thesis was to determine how different module types compared to
each other under elevated conditions namely in Barbados and other Caribbean countries which experience
41
elevated temperatures year round. When comparing modules, the comparison or evaluation could be done on
several levels; 1) from a purely performance standpoint, 2) from a purely cost perspective or 3) from a financial
performance level which takes both cost and energy production into consideration. All 3 levels of assessments
were done in this thesis. As expected, from Figure 2, the FirstSolar CdTe had the highest specific yield and
performance ratio out of the rest of module types. This is in part due to its high band gap (direct) in comparison
to crystalline modules. Sanyo HIT modules also had a higher performance ratio and specific yield compared to
the crystalline modules due to its layer of amorphous silicon which also has a higher band gap (direct) than
typical crystalline silicon which has an indirect band gap. This increased performance for CdTe modules is not
justifiable on rooftops due to the modules significantly lower efficiency and higher cost as compared to
polycrystalline modules. The increase in performance in HIT modules over polycrystalline modules cannot be
justified due to the significant increase in cost of these modules over multicrystaline. In general unless space is a
major constraint and the user requires a certain output then at the current price these HIT modules are not
worth the investment.
In this study it must be noted the difference between the financial projections in the preliminary design and
detail design. The detailed design requires a significant increase in work, analysis and time yet yield very similar
financial metrics compared to the preliminary design. The preliminary design for the Trina system yielded an
LOCE of $0.210/kWh whereas for the final designed system the LOCE was $0.201/kWh. Similarly the ROI for
the preliminary design was 16% whereas for the final design it was found to be 17%. This highlights the ability
to provide realistic financial metrics to a client without detailed in depth analysis which requires more resources
and time.
Another discussion point is the fact that once there is no major shading that it is beneficial to place modules on
every section of the roof as was reflected in this study whereby the worst orientation and tilt (Azimuth 144o, tilt
15o) only gave a 6.2% loss with respect to irradiation on a perfectly oriented and tilted surface in Barbados
(Azimuth 0o, tilt 12o). This statement is justified by the high annual irradiation on this surface of 1843kWh/m2.
This rationale can be extended throughout the Caribbean islands on roofs with relatively low tilts (<15o) due to
the high annual irradiation the region receives year round and the sun’s relatively high position in the sky year
round.
Another result from this study which can be extended to the eastern island chain down to Trinidad and Tobago,
Curacao and Aruba would be the design of a racking system to withstand 130mph based on Figure 8. The
racking systems designed for these locations will generally be required to have shorter rail spans and cantilever
distances than in areas which are not prone to hurricane force winds.
At the outset of this study it was aimed to show that a commercially installed PV system in Barbados would
result in a substantial savings in electricity cost to the customer. It was also aimed to show that a well-designed
system could be achieved with good projected performance at a globally cost competitive price ($2.59/Wp).
A simplification made during the comparison of different systems was modeling them based on one roof tilt and
orientation and not modeling them exactly how the final simulation was done. The reason for this is that the
final simulation required 11 simulations in PVSYT and to do this for all 4 systems would require 44 PVSYST
simulations to be made. If time were not a constraint this could be done. However based on the initial
simulation done for Trina and the Final simulation done for the same modules, the difference in Performance
Ratios and Specific Outputs was not major.
An assumption was made that there is room in the main electrical panel for the feed in over the proposed PV
system. It is assumed that the man electrical panel is rated to with stand the addition of a 600A breaker required
by the system. In reality if the project proceeds, this will have to be verified. In the event that the Main Panel
cannot handle the addition of this electrical source then the building switch gear will have to be upgraded.
Based on the restriction of the FIT size for commercial systems being 150KW, it is assumed that the net
metered configuration will always produce less than the base load requirements of Carlton thereby not exceeding
42
this Feed In limit set by the utility. Future work would be to accurately determine hourly loads to ensure that
this feed in limit set by the utility is not breeched.
An assumption of a project of 20 years after which the system yields no more return to the user is conservative
as the system could well continue to function and generate electricity up to 25 years and beyond.
One of the major limitations of this study and in Barbados is the lack of a long term guaranteed FIT rate from
the utility for clients wishing to feed clean energy to the grid. This makes it difficult for potential customers to
have an idea of their investment 20years down the line with no guarantees on a fixed rate from the utility above
2 years currently. This dilemma is circumvented through the analysis of savings to the user at the cost of
electricity rather than a FIT rate as was done in this study. The conservative financial projects at a fixed cost of
electricity to the user at $0.35/kWh act as a bounding analysis for any countries within the region whose cost of
electricity is similar. The analysis showing a LOCE of $0.201/kWh illustrates the merit of a commercial system
in other islands where the cost of electricity to user is greater than $0.201/kWh. A sensitivity analysis for
different electricity rates as well as for different discount rates would allow for a more comprehensive
understanding of the benefits of a PV system in different islands where the financial interest rates and cost of
electricity vary compared to Barbados.
A limitation of this thesis is that a structural analysis has not been done and it is assumed that the structural
integrity of Carlton’s roof is sound. However a full structural analysis has to be done before the project can be
given the go ahead. It is assumed in this study that at no point during the life time of the project that the existing
roof will have to be replaced. The real life implications of a re-roof to the user will be costly as the modules will
have to be removed then reinstalled over the re-roof period all while electricity production will be halted. A way
of minimizing these effects would be to re-roof the building in stages.
Another assumption was that total capital cost for the PV System was through equity. A sensitivity analysis on
various loan fractions at different interest rates would give the client a good picture of his or her best investment
option. It would be beneficial to do a financial analysis based on the assumption that all proposed incentives are
legislated instead of just the conservative approach adopted in this thesis. An example of this would be modeling
the effect of a soft loan whereby 150% of the interest is covered by the Government of Barbados for
investment in renewable energy systems (Nation News Barbados, 2012).
Future work would be to determine how Carlton’s load can be further offset from 28% to above 50% through
the utilization of available parking lot area to create parking lot arrays and shade structures for cars. In addition
to this, it would be useful to collect hourly load data from the meters on site to have a clearer picture of
electricity usage throughout the day. Ideally the hourly system production would be superimposed on this to
give a visual of how much electricity is being offset by the PV system on an hourly basis. This information could
then be used to further try to shift some of the user’s loads and reduced electricity usage to work in tandem with
the production curve of the PV system.
It would be beneficial to do a PVSYST system analysis for the same final system size however substituting string
inverters for multiple central inverters to see their effect on total system cost and performance. Lastly, it would
be ideal to repeat the PVSYST analysis using monocrystaline modules to confirm previous assessments that the
marginal increase in performance over polycrystalline modules isn’t warranted by the increased costs of these
modules.
7 Conclusion
The final financial figures for the fully designed 425kWp system provide compelling grounds for the investment
in this initiative from a client and investors perspective. A simple payback period of 5.4 years and a Return on
Investment of 17%, a NPV of $564,000 and LCOE of $0.20/kWh make the feasibility of this project unignorable. Further-more when an escalating cost of electricity of 3% per annum is considered the Return on
Investment jumps to 20% with a cumulative cash flow of USD $4,076,000 over 20 years. On top of this, when
43
the proposed Government incentives listed in appendix become legislated this shall further bolster the
attractiveness of commercial PV systems being installed not only on Carlton but island wide as well. These
financial metrics are even supported still by the cost of electricity over the system life being $0.20 USD/kWh,
well below grid parity of the $0.35USD /kWh being paid at the commercial level (which includes connection
cost and VAT). As the price of PV systems decline and the cost of electricity inevitably rises, this gap in the cost
of electricity generated by PV compared to that paid to utility will only continue to increase.
Even after covering sub optimally oriented and tilted roofs with PV, sound technical performance numbers
emerge with a specific yields of 1467 kWh/kWp and a PR of 0.77 which cannot be replicated by many other
geographical locations and further adds to why PV makes absolute sense in Barbados with or without incentives.
Due to existing legislation of a maximum 150kW feed in limit to the grid (Williams, 2013) for commercial
generating systems, the proposed system for Carlton (425kW) has to be configured with net production to grid
which effectively acts as a net metering installation as it unlikely that production from the generation system will
exceed demand even during closed days because of high base load energy requirements intrinsic to supermarkets
which require round the clock refrigeration and cooling. It is hoped that this feed in limit is increased and an
attractive feed in rate above the current $0.36 USD/kWh is guaranteed for a substantial period of time to
further encourage investment and installation of PV Systems Island wide.
While it is observed that greater system losses are incurred due to high year round module operating
temperatures, these losses are more than compensated for with a higher annual irradiation 1951 kWh/m2 on a
flat surface (PVSYST V5.59, 2012), and an average of 5.7 peak sun hours per day (Rogers et al. 2011) compared
to other less favorable regions in which PV is flourishing.
Due to CdTe’s superior performance under higher temperature, it becomes a candidate for ground mount
systems where space is not as big of a constraint as is typical for commercial roof top applications. If the price
per watt were to drop from $0.95/W approaching that of polycrystalline then it would be a serious candidate for
customers who are not looking to maximize system production but merely looking to offset cost for social and
environmental reasons. Similarly, Sanyo HIT cost of $1.35/W cannot be justified for its higher output per m2
and higher specific yields from a financial perspective. A financial rate of return of 12% fails in comparison to
that of Trina and First Solar systems financially speaking. By varying system cost per watt prices in Resco’s
financial model, it was found that a module cost per watt of below $1.00/W would have to be made before
these HIT modules are considered as a competitive solution for maximizing a client’s return on investment.
The importance of detailed design and following the design steps from feasibility study all the way through to
accurate BOS take offs is highlighted by the complexity of the roof of Carlton with seven different roof sections
with varied tilt and orientations. Selection of an appropriate inverter solution is as fundamental as selecting the
correct module type for the given application. An accurate shading analysis acts to influence system size and
how the strings are laid out. A detailed array layout, MID, SLD and conductor schedule are all crucial for
installers, electricians engineers as well as the Utility to effectively understand and build a successful PV system.
Future considerations would be for Carlton to further develop its PV generating capacity by utilizing the
extensive existing car park area in the form of solar parking lots in efforts to further offset electricity cost from
BL&P which will continue to rise.
The two main limitations encountered in this project were that PVSYST is limited to only modeling two
orientations at a time while providing linear shading simulations accounting for one orientation. This increases
the number of simulations which were required to get a system level impact on shading as well as decreases the
accuracy of the simulated losses due to shading. In addition, it was difficult to get the appropriate parties
involved for a complete structural evaluation due to the inability of directly engaging the appropriate parties as
well as due to the lack of remuneration to be provided for their work. However, when this project comes to
fruition, this cost will be embedded in the project cost.
44
In closing, a commercial PV system installed on Carlton provides a lucrative return on investment, cuts down on
fossil fuel consumption and at 1700 kWh/ barrel of oil (Silverman, n.d), diverts approximately 367 barrels of oil
annually. At 317kg of CO2 per barrel of oil (Monibot, 2009), this represents a diversion of 116 tonnes of CO2
annually which in turn equates to a savings of 6840 barrels of oil over 20 years or 2,168 tonnes of CO2 in that
same time period. This shows the
social and environmental benefit of a commercial PV system in helping Barbados reduce its carbon foot print in
addition to the amount of foreign exchange spent on oil imports annually. All the above factors are compelling
evidence that Barbados can achieve high levels of PV penetration on a commercial scale.
45
8 References
ADRANI.K, BARBOSE.G, MARGOLIS.R, WISER.R, FELDMAN.D & ONG.S.(2012) Benchmarking NonHardware Balance of System (Soft) Costs for U.S Photovoltaic Systems Using a Data-Driven Analysis from PV
Installer Survey Results. National Renewable Energy Laboratory (NREL)
CHEN, J. (2007) European Solar Energy Engineering School Master Thesis No. 87: Design of 2.2MW Solar PV
roof in Spain.
DURSCHER.C, FALK.A, REMMERS.K (2007) Photovoltaics for Professionals: Solar Electric Systems
Marketing Design & Installation.
ELERCHIMAN.R (2013) Utilizing String Inverters in Large Commercial Systems. Available from:
http://solarprofessional.com/ Issue 6.5[Accessed 10/07/13]
GIBBS.T (2008) Caribbean Application Document for ASCE7-05 Chapter 6 Wind Loads with references to
Chapter 1 General and 2 Combinations of Loads.
HERMES.M (2013) SMA German Solar Energy Forum Crowne Plaza Manila. Available from:
http://gpcci.org/home/wp-content/uploads/2013/05/10-SMA-presentation-Forum-May-2013.pdf [Accessed
15/08/13]
HULD.T, GOTTSCHALG.R, BEYER.H, TOPIC.M (2012), Mapping the performance of PV modules, effects
of module type and data averaging.
Ministry of Finance and Economic Affairs (2012) Barbados Economic and Social Report 2012. Available from:
http://www.economicaffairs.gov.bb/archive-detail.php?id=324 [Accessed 23/08/13]
MONIBOT.G (2009) Not Even Wrong. Available from: http://www.monbiot.com/2009/08/31/not-evenwrong/ [Accessed 25/08/13]
Nation News Barbados (2013) Barbados 2012 Financial Statement and Budgetary Proposal (Part 3). Available
from http://www.nationnews.com/articles/view/2012-financial-statement-and-budgetary-proposals-part-3/
[Accessed 08/08/13]
NFPA 70 (2010) National Electrical Code 2011 Edition
PVSYST V5.59 (2012) Photovoltaic Software. Available from http://www.pvsyst.com/en/
ROGERS, T., CHMUTINA, K. & MOSELEY, L. (2011) The potential of PV installations in SIDS-an example
in the island of Barbados. University of the West Indies.
SILVERMAN.D (n.d) Energy Units and Conversions. Available from: University of California Irvine- Physics
and Astrononmy Department http://www.physics.uci.edu/~silverma/units.html
[Accessed 25/08/13]
SMITH.S, SHIAO.M.(2012) Solar PV Balance of System Markets: Technologies, Cost and Leading Companies,
2013-2016. Green Tech Medial Research
Southwire Company (2005)www.Southwire.com, Voltage Drop Calculator. Available
from:http://www.liteco.ca/Portals/_Rainbow/images/default/Tools/Southwire/voltagedropcalculator.htm
[Accessed 15/07/13]
46
STREVEL, N., TRIPPEL L., & GLOECKLER M. (2012) Performance characterization and superior energy
yield of First Solar PV power plants in high-temperature conditions. Available 17th Edition of Photovoltaics
International journal, August 2012.
The Barbados Advocate (2013), Green Energy Push, 07-30-2013. Available from
http://www.barbadosadvocate.com/newsitem.asp?more=local&NewsID=31711 [Accessed 09/08/13]
The Barbados Light & Power, Renewable Energy Rider. Available from
http://www.blpc.com.bb/search_details.cfm?Item_ID=75#A [Accessed 09/08/13]
The Barbados Light & Power Company Limited (2011) Requirements for Grid Interconnection of Renewable
Generation Systems.
WILLIAMS.P (2013) Renewable Energy Developments in Barbados, Business Barbados. Available from
http://businessbarbados.com/industries/energy-and-utilities/renewable-energy-developments-barbados/
[Accessed 08/08/13]
ZIPP. T (2013) Top 10 Solar PV Module Suppliers of 2012. Available from:
http://www.solarpowerworldonline.com/2013/04/top-10-solar-pv-module-suppliers-of-2012/ [Accessed
23/08/13]
List of Appendices
Appendix A: List of Exisiting PV Installations in Barbados ............................................. 48
Appendix B NEC Code..................................................................................................... 49
Appendix C: Application for Grid Interconnection .......................................................... 50
Appendix D: Barbados Historical Temperature Trends .................................................. 51
Appendix E: Inverter Input Details for Sanyo, Ying Li & First Solar Modules ................ 52
Appendix F: Detailed Project Cost for proposed 425kW PV System at Carlton ............... 54
Appendix G: Barbados Utility Production & Consumption 2009-2012 ............................. 56
Appendix H: Proposed Government Incentives for Renewable Energy Sector .............. 57
Appendix I: Module Component Specification Sheets ..................................................... 59
Appendix J: Inverter Specification Sheet ........................................................................... 67
Appendix K: DPW Racking Specification Sheet ............................................................... 70
Appendix L: Trina Installation Manual ............................................................................ 76
Appendix M: Sample from preliminary PVSYST simulations .......................................... 83
Appendix N: Sample from final PVYSYT simulations ...................................................... 86
Appendix O: Carlton A-1 Base Building Layout ............................................................ 106
Appendix P: Carlton A-1 Array Layout ............................................................................ 107
Appendix Q: Carlton A-1 Site Plan.................................................................................. 108
Appendix R: Carlton A-1 MID Sheet 1 ............................................................................ 109
Appendix S: Carlton A-1 SLD ........................................................................................... 111
Appendix T: Carlton A-1 Conductor Schedule ................................................................. 112
47
Appendices
Appendix A: List of Exisiting PV Installations in Barbados
Figure 23 List of existing PV installations in Barbados (SMA Portal, 2012)
48
Appendix B NEC Code
49
Appendix C: Application for Grid Interconnection
50
Appendix D: Barbados Historical Temperature Trends
Appendix D-1: Figure showing Maximum & Minimum Temperatures for Bridgetown Barbados during 2012
Source:www.weatherspark.com
Appendix D-2: Maximum recorded monthly temperature in Bridgetown Barbados over the past 20 years
Source:www.weatherbase.com
51
Appendix E: Inverter Input Details for Sanyo, Ying Li & First Solar
Modules
Inverter Input Details for Sanyo HIT-240HDE4
Inverter Type
Sunny Tripower STP 15000TL
Mpp voltage Range
360-800V
String Length
15
Strings Per MPP input A
5
Strings Per MPP input B
1
Total # of modules Per 15kW inverter
90
Inverter Details
Input A
Input B
Module Count
75
15
Max Inverter Vmpp
Max input Voltage, Voc @18oC
Min String Voltage to be within
Inverter Vmpp
Min Input Voltage Vmpp @ 85oC
Maximum Input Current
String Input Current (Impp)
Maximum electronic Fuse Short
Circuit Current
Inverter Short Circuit Current
800V
665V
800V
665V
360V
434V
33A
27.55A
360V
434V
11A
5.51A
40A
36.85A
12.5A
7.37A
Inverter Input Details for Ying Li YL280-35b
Inverter Type
Sunny Tripower STP 15000TL
Mpp voltage Range
360-800V
String Length
15
Strings Per MPP input A
4
Strings Per MPP input B
1
Total # of modules Per 15kW inverter
Inverter Details
Module Count
Max allowable input Inverter Vmpp
Voltage
Max input Voltage, Voc @18oC
Min String Voltage to be within
Inverter Vmpp
Min Input Voltage Vmpp @ 85oC
Maximum Input Current
String Input Current (Impp)
Maximum electronic Fuse Short
Circuit Current
Inverter Short Circuit Current (Isc)
75
52
Input A
60
Input B
15
800V
693V
800V
693V
360V
414V
33A
31.56
360V
414V
11A
7.89A
40A
33.4A
12.5A
8.35A
Inverter Input Details for First Solar FS-390
Inverter Type
Sunny Tripower STP 15000TL
Mpp voltage Range
360-800V
10
String Length
Strings Per MPP input A
5
Strings Per MPP input B
1
4 inputs with 4
parallel strings, 1 with
Number of Parallel Strings to fit in Input
A
17
one string
Number of Parallel Strings to fit in Input
1 input with 5 parallel
B
strings
5
Inverter Details
Input A
Input B
Module Count
170
50
Max allowable input Voltage per
Inverter Vmpp
800V
800V
Max input Voltage, Voc @18oC
614V
614V
Min String Voltage to be within Inverter
Vmpp
360V
360V
Min Input Voltage Vmpp @ 85oC
417V
417V
Maximum Input Current
33A
11A
String Input Current (Impp)
32.3A
9.5A
Maximum electronic Fuse Short Circuit
Current
40A
12.5A
Inverter Short Circuit Current (Isc)
35A
10.3A
53
Appendix F: Detailed Project Cost for proposed 425kW PV System at Carlton
Table 15 Detailed Bill of Material Cost for Carlton
Material Costing for Carlton A-1 - Barbados
EXTENSION
Description
4" CONDUIT - EMT
4" ELBOW 90 DEG - EMT
4" COUPLING SS DC - EMT
4" CONN SS STL - EMT
4" COUPLING SS STL - EMT
4" CONDUIT - RMC - GALV
4" COUPLING - RMC - GALV
4" ELBOW 90 DEG - RMC - GALV
4" CONN THRD HUB INSUL - RMC - MALL OR STL
4" LOCKNUT
4" LOCKNUT GRDG - MALL
4" BUSHING - PLASTIC
4" MEASURE CUT & THREAD LABOR - RMC - GALV
4" EMT-GRC STRUT CLAMP
4" GRC-EMT STRUT CLAMP
1/2" PVC J-BOX O-RING
1" PVC J-BOX O-RING
#350MCM RW90
#500MCM RW90
#2 RW90 GREEN
#10/3C TECK 1KV AL/PVC
#4/3C TECK 1KV AL/PVC
#3/3C TECK 1KV AL/PVC
#6/3C TECK 1KV STL/PVC
#10466 CONN (1/2") WT
#10469 CONN (1") WT
#2 HYPRESS 1-HOLE LUG
350MCM HYPRESS 1-HOLE LUG
500MCM HYPRESS 1-HOLE LUG
18" AL LADDER TRAY 6" RUNGS 3" DEEP
18" AL LADDER HORZ 90 ELBOW
3/8x 3 SLEEVE ANCHOR W/ HEX NUT - 1 1/4" MIN DEPTH
3/16" FLAT WASHER (PLTD)
#10x1-1/2" SELF TAPPING SCREW
TY28MX 14.19 BLK CABLE TIES
600A-51" BRKR HEIGHT DPB
30A 3P BREAKER BOLT-ON
600A 250V FUSIBLE NEMA 3R/12
600A 600V NON-FUS NEMA 3R
DSN 600A NEUTRAL BAR
600A 600V TIME DELAY FUSE
48x48x12 METER CABINET
TRINA 250W
POWER-FAB-CRS
DURABLOCK
MC4 SOLAR CON MALE 3PIECE SET 10AWG
MC4 SOLAR CON FEMALE 3PIECE SET 10AWG
#10 RPVU90
SUNNY TRIPOWER 15000TL
Inverter Shade Structure
Webbox System monitoring
MISC MATERIALS
Totals
Quantity Price Unit
40
C
8
C
2
C
8
C
6
C
120
C
2
C
8
C
8
C
16
C
4
E
16
C
4
C
4
C
7
C
6
C
34
C
552
M
380
M
176
M
188
M
1706
M
2056
M
650
M
6
E
34
E
6
C
8
C
6
C
80
E
6
E
18
C
640
C
640
C
881
C
1
E
20
E
1
E
1
E
2
E
3
E
1
E
1700
E
1700
E
320
E
150
E
150
E
37000
M
20
E
20
E
1
E
1
E
Net Cost
$ 540.27
$ 4,830.28
$ 1,049.54
$ 798.90
$ 540.67
$ 1,287.71
$ 6,919.44
$15,326.94
$12,376.80
$ 168.89
$
39.49
$
85.11
$
$ 533.80
$ 533.80
$
22.50
$
28.75
$ 6,972.69
$ 9,501.26
$ 1,472.20
$ 1,600.38
$ 4,920.48
$ 4,925.10
$ 3,838.89
$
15.27
$
27.03
$ 774.01
$ 2,275.88
$ 4,379.14
$
85.00
$
65.00
$
63.84
$
36.00
$
6.42
$
10.41
$ 5,000.00
$ 149.54
$ 1,570.59
$ 2,043.56
$
79.68
$ 180.93
$ 351.00
$ 195.00
$
50.00
$
6.10
$
2.85
$
2.79
$ 314.00
$ 5,057.08
$ 515.00
$ 1,548.00
$15,000.00
54
Labor Labor Unit
13.7
C
80
C
17
C
36
C
17
C
23.4
C
0
C
186
C
72
C
15
C
20
C
9
C
77
C
28
C
28
C
10
C
10
C
31.5
M
38.7
M
12.42
M
21.15
M
45.25
M
50.2
M
37.8
M
35
C
50
C
22.5
C
52.2
C
52.2
C
0.15
E
0.9
E
12
C
1
C
3
C
6
C
18
E
0.4
E
6
E
6
E
0.5
E
0.18
E
3
E
0.42
E
0.67
E
0.05
E
0.06
E
0.06
E
4.32
M
2
E
2
E
7
E
Total Material Total Hours
$
216.11
5.48
$
386.42
6.4
$
20.99
0.34
$
63.91
2.88
$
32.44
1.02
$
1,545.25
28.08
$
138.39
0
$
1,226.16
14.88
$
990.14
5.76
$
27.02
2.4
$
157.96
0.8
$
13.62
1.44
$
3.08
$
21.35
1.12
$
37.37
1.96
$
1.35
0.6
$
9.78
3.4
$
3,848.92
17.388
$
3,610.48
14.706
$
259.11 2.18592
$
300.87
3.9762
$
8,394.34 77.1965
$ 10,126.01 103.2112
$
2,495.28
24.57
$
91.62
2.1
$
919.02
17
$
46.44
1.35
$
182.07
4.176
$
262.75
3.132
$
6,800.00
12
$
390.00
5.4
$
11.49
2.16
$
230.40
6.4
$
41.09
19.2
$
91.71
52.86
$
5,000.00
18
$
2,990.80
8
$
1,570.59
6
$
2,043.56
6
$
159.36
1
$
542.79
0.54
$
351.00
3
$ 331,500.00
714
$ 85,000.00
1139
$
1,952.00
16
$
427.50
9
$
418.50
9
$ 11,618.00
159.84
$ 101,141.60
40
$ 10,300.00
40
$
1,548.00
7
$ 15,000.00
0
$ 614,553.55 2625.03
Table 16 Direct Labor breakdown for Carlton A-1
DIRECT LABOR
Labor Type
FOREMAN
JOURNEYMAN
APPRENTICE
LABOURER
Totals
Crew Hours
Rate $
1 201.926 $ 41.00
2 403.851 $ 41.00
4 807.702 $ 30.00
6 1211.55 $ 21.00
13 2,625.03 $ 28.38
Total
$ 8,278.95
$ 16,557.89
$ 24,231.06
$ 25,442.62
$ 74,510.52
Table 17 Incidental Labour metrics for Carlton A-1
INCIDENTAL LABOUR
Incidental Labor
JOB MOBILIZATION
SUPERVISOR
TRUCKING
FOREMAN TRAVEL
TIME
SITE SURVEY
MISC SHOP
SAFETY
FOREMAN MISC
COMMISSIONING
MONITORING
CRANING
Totals
Hours Rate $
64 $ 28.38
60 $ 41.00
30 $ 50.00
50
16
70
360
20
85
32
28
787
$
$
$
$
$
$
$
$
$
35.00
32.00
28.38
28.38
35.00
38.00
38.00
200.00
32.26
Total
$ 1,816.32
$ 2,460.00
$ 1,500.00
$
$
$
$
$
$
$
$
$
1,750.00
512.00
1,986.60
10,216.80
700.00
3,230.00
1,216.00
5,600.00
30,987.72
Table 18 General Expenses for Carlton A-1
GENERAL EXPENSES
General Expenses
LAMACOIDS
SITE OFFICE
TOILETS
SEA CANS
STAIR TOWER
INSURANCE
PERMITTING
MISC
Totals
Quantity Cost/Unit
1 $ 750.00
1 $ 1,583.00
1 $ 673.00
1 $ 390.00
1 $ 2,542.00
1 $ 4,500.00
1 $ 4,500.00
1 $ 750.00
55
Total Cost
$
750.00
$ 1,583.00
$
673.00
$
390.00
$ 2,542.00
$ 4,500.00
$ 4,500.00
$
750.00
$ 15,688.00
Appendix G: Barbados Utility Production & Consumption 2009-2012
56
Appendix H: Proposed Government Incentives for Renewable Energy
Sector
INCENTIVES FOR BUSINESSES GENERATING and DISTRIBUTING ELECTRICITY FROM A
RENEWABLE ENERGY SOURCE (RE), BUSINESSES PRODUCING, DISTRIBUTING and/or
INSTALLING RENEWABLE ENERGY SYSTEMS FOR ELECTRICITY GENERATION and
ENERGY EFFICIENT (EE) PRODUCTS(Nation News Barbados, 2013)
• Government will provide financial assistance of $100 million in the form of low-interest loans over an
eight (8) year period to tool and capitalize the industry, and to provide improved credit to the consumers
of RE and EE products produced locally. This loan facility will be available to the industry via the
financial intermediaries (including banks, credit unions, and finance companies) and administered by way
of a special window under the Industrial Credit Fund and the Credit Guarantee Scheme of the Central
Bank of Barbados;
• Building materials and supplies for construction of a facility dedicated to the generation and sale of
electricity from a renewable source shall be duty free and VAT free;
• A zero rate of VAT will be applied to all RE and EE systems and products produced in Barbados;
• Developers, manufacturers and installers of RE products will be granted an income tax holiday of 10
years;
• Eligible businesses will now have a 150% deductible on interest paid on loans entered into for:
• upgrading an existing property so as to generate and sell electricity from a renewable source;
• Construction of a new facility to generate and sell electricity from a renewable source;
• Construction of a facility for the installation or supply of renewable energy generation systems or
energy efficient products ;
• Eligible businesses will now have a 150% deductible on expenditures for staff training related directly to
the generation and sale of electricity from a renewable source in the installation, as well as the installation
and servicing of renewable energy electricity generation systems or energy efficient products;
• Individuals will be able to claim the funds spent on RE and/or EE training provided by educational and
vocational institutions that are approved by the Barbados Accreditation Council as a separate deduction
on their income tax. This deduction can also be made by the parents of minors and adult students (up to
the age of 25 years) who are not working and who are studying in the area of RE and/or EE;
• Eligible businesses will now have a 150% deductible on expenditures directly related to the marketing of
products for the generation and sale of electricity from a renewable source in the installation, as well as
the installation and servicing of renewable energy electricity generation systems or energy efficient
products;
• Eligible businesses will now have a 150% deductible on expenditures for product development/research
related directly to the generation and sale of electricity from a renewable source in the installation, as well
as the installation and servicing of renewable energy electricity generation systems or energy efficient
products;
57
INCENTIVES FOR INVESTORS IN BUSINESSES GENERATING AND DISTRIBUTING
ELECTRICITY FROM A RENEWABLE ENERGY SOURCE (RE), BUSINESSES PRODUCING,
DISTRIBUTING AND/OR INSTALLING RENEWABLE ENERGY SYSTEMS FOR
ELECTRICITY GENERATION and ENERGY EFFICIENT (EE) PRODUCTS (Nation News
Barbados 2013):
•
•
•
•
•
Interest earned by financial intermediaries (including banks, credit unions, and Finance companies) for
financing the development, manufacturing, and installations of RE and EE products and technologies
shall be free of withholding taxes for 10 years;
Venture capital funds where investments are made in the RE and EE sectors shall be free from
corporation taxes for a period of 10 years;
Contributions to venture capital where investments are directed to the RE and EE sectors shall be
deductible from taxable income for a period of 10 years;
Dividends for shareholders of companies investing in entrepreneurial businesses engaged in installation
or supply of renewable energy electricity generation systems or energy efficient products shall be
exempted from withholding taxes for a 10 year period
The income households earn from the sale of electricity produced from the utilization of RE equipment
shall be exempted from all taxes.
58
Appendix I: Module Component Specification Sheets
59
60
61
62
63
64
65
66
Appendix J: Inverter Specification Sheet
67
68
69
Appendix K: DPW Racking Specification Sheet
70
71
72
73
74
75
Appendix L: Trina Installation Manual
76
77
78
79
80
81
82
Appendix M: Sample from preliminary PVSYST simulations
83
84
85
Appendix N: Sample from final PVYSYT simulations
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
Appendix O: Carlton A-1 Base Building Layout
106
Appendix P: Carlton A-1 Array Layout
107
Appendix Q: Carlton A-1 Site Plan
108
Appendix R: Carlton A-1 MID Sheet 1
109
Appendix R: Carlton A-1 MID Sheet 2
110
Appendix S: Carlton A-1 SLD
111
Appendix T: Carlton A-1 Conductor Schedule
112
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement