DistributionCategory UC-63a
SAND81-7182Unlimited Release
Printed March 1982
DETAILED RESIDENTIALPHOTOVOLTAIC SYSTEM
REFERENCE DESIGNS
FINAL REPORT
E. M. Mehalick, R. Landes, N. TruncellitoGeneral Electric Energy Systems and Technology Division
King of Prussia, PA 19406
ABSTRACT
This document summarizes the detailed residentialphotovoltaic system designs developed by General Electricfor Sandia National Laboratories. The specific designsare presented in SAND79-7056, SAND80-7148, SAND80-7170,SAND80-7171, SAND80-7172, and SAND80-7173.
Prepared for Sandia National Laboratories under Contract 13-8779.
"\I
Section
TABLE OF CONTENTS
Page
1
2
3
INTRODUCTION • •
ObjectiveBackground •
SYSTEM SELECTION •
Classification and Evaluation of PV SystemsClassification of PV Systems • • • • • • • •Evaluation of PV Systems Within a Category ••Relative Effectiveness By Category (Collector)System/Site Selections. • •••••••
SYSTEM DESIGN SUMMARIES
Type
1-1
1-11-2
2-1
2"-22-22-32-52-6
3-1
A PV System for An All-Electric Residence in the Southwest 3-4Side-By-Side PV/Thermal System for the Northeast ••••• 3-14A PV System With Battery Storage for the Southwest. • • • • • 3-27Passive House Design for the Northeast • • • • • • • • • • •• 3-37Integral Mounted PV Array for the Southeast • • 3-46A PV System for a Temperate Climate •• ". • • 3-55
4
5
DESIGN CONCERNS
Array Sizing • • • • • •Roof Constraints • • • • • • • • • •Array Mounting Approach • • • •Power Conversion Subsystem • • • • • • • • •Isolation Transformer and Grounding • • • • • •Exterior Disconnect SwitchesBattery Location Concerns ••••Module Interconnection • •• • • • • • • • •Fire Safety
DESIGN STUDIES •
4-1
4-14-14-24-24-34-34-44-44-4
5-1
Power Conversion Subsystem Correlations • • • • • • •• 5-1Sensitivity of Photovoltaic Module Performance to RoofInsulation for the Integral Mount Configuration • • • • • 5-8
Alternate Battery System Shunt Analysis • • • • • 5-10System Current Characteristics • • • • • • • • • • • •• 5-15Array Open Circuit Voltage. • •• • • • • • • • •• 5-19
6
APPENDIX A
REFERENCES
SUMMARY OF SYSTEM CONFIGURATION EVALUATIONS
6-1
A-I
iii
Figure
LIST OF ILLUSTRATIONS
Page
1-11-21-31-4
2-1
Utility Feedback System ••••On-Site Battery Storage System ••••Side-By-Side PV/Thermal System. • •••Alternate PV/Thermal System Configurations
Alternate Mounting Techniques for PV Modules •
1-2. . . .. 1-3
1-31-4
• • • • 2-7
5-1
5-25-35-45-55-65-75-8
5-9
5-105-115-12
5-13
5-14
5-15
5-165-175-18
5-19
5-20
iv
System Losses for the Southeast Design as a Functionof Limiting Voltage Range. • • • • • • •••
System Losses for Boston"Design 1 •System Losses for Phoenix, Design 1System Losses for Boston, Design 4 . . . . . . .System Losses for Santa Maria, Design 6 •••.System Losses for an El Paso Single Family ResidenceComparison Between Minimum Loss Voltage and Insolation.Correlation Between Minimum Loss Voltage and NormalizedAmbient Temperature • • • • • • • • • • • • • • • • • •
Correlation Between Normalized %Increase in Loss WithNormalized Voltage Operating Range About Minimum Loss Voltage
Module Cross Section for Integral Mount Configuration • • • •Assumed Attic Temperature as a Function of Ambient TemperatureCell Temperature and Module Efficiency as a Function ofAmbient Conditions for the Integral Mount Configuration.
Cell Temperature and Module Efficiency as a Function ofAttic Temperature for the Integral Mount Configuration
Available Power, Hot Water Usage, and Preheat TankTemperature for a PV System in Albuquerque
Current Duration for the Sixth DesignSanta Maria, CA • • • • • • • • • • •
Current Distribution for the Fourth Design, Boston • • • • • •Current Duration for the Fourth Design PV Array, Madison.Current Distribution as a Function of Solar Array Energyfor Boston . . • • . . • . . . . . . . . • . . . . . .
Current Distribution as a Function of Solar Array Energy,Madison .• G • • • • • • • • • • • • • • • • •
Effects of Ambient Temperature and Solar Intensity on OpenCircuit Voltage. • • • • • • • ••••••••
5-25-35-35-45-45-55-5
5-7
5-75-95-11
5-11
5-12
5-13
5-165-165-17
5-17
5-18
5-19
Table
2-12-22-32-42-5
LIST OF TABLES
Page
Summary of System Configurations • • • • • • • • • • • • • •• 2-2Generic System Configurations for Residential Demands. • 2-4Summary of System Options ••••••••••••••• 2-4Selected System Designs •• • • • • • • • • • • • • • • 2-8Summary of Subsystem Options Addressed in the Designs 2-8
Economic Assumptions for Optimization • • • • • • • • •3-1
5-1 Material Combinations for Thermal Path from Solar Cellto Attic . . . . . . . . . . . . . . . . . . . . . .
v
. . . .
3-2
5-9
FOREWORD
The Detailed Residential Photovol taic System Reference Designs Study was per
formed by the Advanced Energy Programs Department (AEPD) of the General Electric
Company, Energy Systems and Technology Division under Sandia National Laborato
ries, Contract 13-8779. Mr. E.M. Mehalick served as the GE Program Manager, and
Dr. G. Jones served as the Sandia Technical Monitor.
The project team led by AEPD included Massdesign Architects and Planners, Inc.,
who provided the details of the house design and analysis support related to the
solar array installation and Johnson and Stover, Inc., who provided the installa
tion drawings and specifications for the electrical equipment associated with the
photovoltaic system.
The following individuals supported the program. From AEPD, Mr. G. 0 I Brien, Mr.
R. Schaeffer and Mr. R. Felice provided electrical PV system design support; Mr.
J. Parker and Mr. N. Truncellito provided system performance analysis; and Mr. R.
Landes reviewed and selected system configurations. Mr. G. Tully, President of
Massdesign supplied his direct support on all of the house designs and system
integration. From Johnson and Stover, Mr. G. Johnson provided all of the
electrical system design details. The program success was a direct result of the
dedicated effort of all the team members.
The quality of the reports written during the program were enhanced by the review
and suggestions of Dr. G. Jones at Sandia.
VI
SECTION 1
INTRODUCTION
Objective
The objective of the Detailed Residential Photovoltaic (PV) System Reference
Designs Program was to develop regionally appropriate detailed photovoltaic
system designs covering the major system options for the 1986 time frame in
habitable residences. The initial selection of the systems was based on previqus
work of all contractors in the PV residential area and consistent with the
National PV Residential Program test plans. The designs were prepared for four
regions of the country: the Southwest, the Northeast, the Southeast, and a
temperate climate (e.g., California or Hawaii). A total of thirteen system
configurations/regions were initially identified and six detailed designs
developed. The systems considered various hardware options for the major
subsystems. The output from the program included a separate report for each
design with a complete system description, including design requirements,
functional characteristics and site characteristics; a block diagram; full scale
electrical lirie drawings; thermal d~awings; pictorial layouts; a summary of
performance characteristics and tradeoffs; and subsystem and component
specifications. The reports have sufficient information to obtain detailed cost
data from independent sources for installation of the proposed photovoltaic
systems. The systems developed can be used as reference designs for typical
equipment requirements, system performance estimates, installation details
and system cost estimates.
The six designs completed are listed below:
1. A PV system for an all-electric Southwest residence employing a direct mountarray
2. A PV/thermal side-by-side system for a Northeast residence
3. A PV system with battery storage for the Southwest employing a standoff array
4. A PV system for a passive house design for the Northeast
5. A PV system for a Southeast residence employing an integral mount array
6. A PV System for a temperate climate with either a standoff or an integralmounted array
The designs provided detailed drawings for installation of direct mounted modules
(both shingle and batten types), standoff mounted modules and integral mounted
modules.
1-1
This report provides a summary description of all the designs, a discussion of
the initial system configuration review and selection, a summary of several
designtradeoffs completed during the program and a discussion of design issues
and COne erns •
Background
A residential PV system consists of three primary subsystems: the array subsys
tem, the power processing and control subsystem and the storage subsystem. In the
simplest configuration, the system is grid connected with feedback of excess
energy to the utility and thus eliminating the storage subsystem (Figure 1-1).
The economic viability of this system is dependent on the sellback credit ratio
(a fraction of the normal utility charge to a customer) that the utility is
willing to offer for the feedback energy. Previous studies have indicated that
thi s system shows economic viability with a sellback credit of 50% or more in
most regions of the country, References 1 and 2. The alternate basic configura
tion incorporates battery storage into the design (Figure 1-2). This option adds
flexibility to. the system and delivers more energy directly to the house loads;
however, it is costly, adds control complexity and increases maintenance
requirements.
In the broader sense, photovoltaic systems can also be combined with thermal
systems; either in a side-by-side system configuration (Figure 1-3) or through
the use of a combined PV/thermal collector. The side-by-side system uses separate
UTILITYBACK UP
PV DC!AC GENERALARRAY INVERTER LOADS
MAX. POWERTRACKER
HEATto---PUMP
"'"'-HOT
WATER
Figure 1-1. Utility Feedback System
1-2
UT'L'TYBACKUP
~~
DCIAC GENERALINVERTER LOADS
SHUNT-:..~
~][J-I PUMP
.1 ~,
- -~
HOT-~.. WATER
Figure 1-2. On-Site Battery Storage System
UTILITYBACK UP
fT;;Bi TES
Figure 1-3. Side-By-Side PV/Thenlal System
PV and thermal arrays, both of which are currently available. Combined PV/T
collectors were ultimately dropped from consideration because of poor observed
and projected performance. All of the configurations include the power processing
and control subsystem which is centered around the dc/ac inverter. This subsystem
provides the interface between the dc photovol taic array and the ac house loads
and the utility.
Several alternate system configurations can also be considered for the side-by
side PV/thermal system (Figure 1-4). The first variation only provides supplemen
tal thermal energy to the domestic hot water heater. This thermal system is less
complex than the combined space heating and domestic hot water heating system and
1-3
FOSSIL FIRED HEATING SY TEM
SOLAR HOT WATER HEATING SYSTEMBACK-UP
Figure 1-4. Alternate PV/Thermal System Configurations
shows better economic viability. The second variation considers the side-by-side
PV/thermal system incorporated into a house utilizing fossil energy for space
conditioning. The actual system design remains the same with economics slightly
less attractive than for the all electric house, based on current costs of fossil
fuel s.
Another system configuration variation for any design includes a waste heat
recovery unit for hot water heating. Several manufacturers are currently
marketing these units which recover heat from the air conditioner or heat pump
refrigerant as it leaves the compressor. The self-enclosed units consist of a
heat exchanger, sealed motor pump, temperature sensor for the domestic hot water
tank and a control valve to shut down the fluid flow when the hot water tank is
at an upper temperature limit. The General Electric Hot Water Bank unit was used
in one of the detailed designs. The annual results indicated a savings of 38% of
the hot water requirements. Relatively low installation costs makes this unit an
attractive option for the system.
The designs described in Section 3 address each of these basic and alternate
system configurations.
1-4
SECTION 2
SYSTEM SELECTION
Residential PV system studies were reviewed to select configurations which appear
most viable in 1986 in three specified geographic regions in the United States.
The three regions were the Northeast, encompassing Madison, WI to Boston, MA; the
Southeast covering Charleston, SC to Miami, FL; and the Southwest which includes
the Albuquerque, NM and Phoenix, AZ climates.
Previous studies reviewed included the following:
• MIT/Lincoln Solar Research Facility Description
Arlington Solar Research Facility Description
Hybrid Systems Study
Optimization Hybrid Systems Study
• GE Space Division
Conceptual Design and Analysis Study
Regional Conceptual Design and Analysis Study
Photovoltaic Residential Prototype Definition Study
• Westinghouse
Conceptual Design and Analysis Study
Regional Conceptual Design and Analysis Study
• Aerospace CorporationPhotovoltaic Total Energy Residential Systems Study
• Spectrolab
Photovoltaic Systems Concept Study
• Martin Marietta
Photovoltaic Residential Prototype Definition Study
• MIT/Energy Laboratory
Grid-connected Photov6ltaic Economic Study
Specific document numbers and dates of issue for each of the above indicated
studies are contained in the list of references.
2-1
Classification and Evaluation of PV Systems
Classification of PV Systems
On the basis of collector configuration, each of the PV system concepts reviewed
were classified into one of the following three categories:
• PV only
• Separate PV/Thermal
• Combined PV/Thermal
A basic schematic was prepared for each system along with a brief description of
its configuration and operation. Table 2-1 lists all of the configurations.
Evaluation of each system included a listing of major advantages and disadvan
tages, as well as a set of conclusions based on the results of regional perform
ance and economic analyses conducted on previous studies. Details of this evalua
tion are presented in Appendix A.
Table 2-1. Summary of System Configurations
Category Identifier System Title
PV Only
Systems
Ia
Ib
Ic
Id
All Electric/Battery
All Electric/Feedback
Fossil Heating/Feedback
All Electric/Maximum Power Tracking/Battery
IIa
IIb
IIc
IId
He
IIf
IIg
Separate
PV/Thermal
Systems
Combined
PV/Thermal
Systems
All Electric/Solar Assisted Heat Pump/Feedback
Direct Solar Heating/Feedback
Solar Rankine Driven Heat Pump/Feedback
Solar Absorption Cooling/Feedback
Air Solar Boosted Heat Pump/Battery
Air Solar Assisted Heat Pump/Battery
Solar Boosted Heat Pump/Feedback
IIIa Solar Assisted Heat Pump/Feedback
IIIb Direct Solar Heating/Feedback
IIIc Solar Boosted Heat Pump/Feedback
IIId Stand Alone/Direct Heating/Battery'--- --"'- 1
2-2
The combination of generic system combinations is represented in the matrix shown
in Table 2-2. The PV only system with feedback is the simplest system. The system
complexity increases with the addition of battery storage or solar thermal
options. Any of the blocks in the table represent a possible system configura
tion. Table 2-3 lists ·the subsystem options which can be included under each of
the system configurations. Further options can exist wi thin a subsystem as the
type of array flat plate mounting technique. There are four possible mounting
options: stand-off mounting, direct mounting, integral roof mounting or rack
mounting. These various options were considered in final design selections.
Evaluation of PV Systems Within a Category
The following set of criteria was used to rate each system within each of the PV
system categories:
• Economic/Performance Results
• Current Technology Status
• Projected Status for 1986
• Development Risk
• System Complexity
• Projected System Cost
• Regional Applicability
• Duplication of System Components
A rating designation of either good, fair or poor was then assigned to each
system for each of the above criteria (see Appendix A). Based on these ratings,
the systems were ranked wi thin their category. The results of this ranking are
presented below.
PV ONLY SYSTEMS
Rank System Title
1 Ib, All Electric/Feedback
2 la, All Elect~ic/Battery
3 Ic, Fossil Heating/Feedback
4 Id, All Electric/Maximum Power Tracking/Battery
2-3
Table 2-2. Generic System Configurations For Residential Demands
BASE ELECTRICAL LOAD}DOMESTIC HOT WATER RESIDENTIALSPACE HEATING DEMANDSSPACE COOLING
SOLAR THERMAL OPTIONS------.,.~
INCREASING
COMPLEXITY
PV ONLY SYSTEMS PV ONLY PLU~ PV ONLY PLUS PV ONLY PLUSWITH UTILITY SOLAR THERMAL SOLAR SPACE HEATING SOLAR HEATING,
FEEDBACK DOMESTIC HOT WATER AND DHW COOLING AND DHW(GRID CONNECTED) SYSTEM
----------- ---------- ---------- ----------APPLICABLE TO ALL SIDE-BY-SIDE ARRAYS SIDE-BY-SIDE ARRAYS SIDE-BY-SIDE ARRAYS
REGIONS OR OR TECHNOLOGY JUMPCOMBINED ARRAYS COMBINED ARRAYS FROM HEATING
APPLICABLE TO APPLICABLE TO SYSTEMSALL REGIONS COLD REGIONS APPLICABLE TO ALL
REGIONS
PV ONLY SYSTEMSWITH BATTERY
STORAGE(GRID CONNECTED)
----------APPLICABLE TO ALL
REGIONS
PV ONLY SYSTEMS II
WITH BATTERIES I'(STAND ALONE)
----------~APPLICABLE TO
ALL REGIONS ,
(j >Z l-ll! X< ww ...Ja: 0U :Ez 8
PHOTOVOLTAIC1OPTIONS
Table 2-3. Summary of Subsystem Options
COLLECTOR SUBSYSTEM STORAGE & POWER CONDITIONING
FLAT PLATE ARRAYS - SHINGLE TYPEPLUS SEVERAL FLAT PANELS
SELECTED CONCENTRATOR ARRAYSROOF MOUNTED SINGLE AXIS TRACK WITH
ADJUSTABLE TILT
SIDE-BY-SIDE PV/THERMAL FLAT PANELS
COMBINED FLAT PLATE PV/THERMAL PANELS
HEATING SYSTEMS COOLING SYSTEMS
UTILITY FEEDBACK
BATTERIES
MAX POWER TRACKING
LINE COMMUTATING INVERTERSELF COMMUTATED INVERTER
DOMESTIC HOT WATER
HEAT PUMP
RESISTANCE HEATING
FOSSIL HEATING
SOLAR SUPPLEMENTED
(SEVERAL VARIATIONS)
HEAT PUMP
ABSORPTION CHILLER
VAPOR COMPRESSION
SOLAR SUPPLEMENTED
SOLAR DRIVEN RANKINE
FOSSIL
RESISTIVE HEAT
SOLAR SUPPLEMENTED
2-4
IIa,
IId,
IIc,
IIb,
IIf,
IIg,
IIe,
Rank
1
2
:3
4
5
6
7
SEPARATE PV/THERMAL SYSTEMS
System Title
All Electric/Solar Assisted Heat Pump/Feedback
Solar Absorption Cooling/Feedback
Solar Rankine Driven Heat Pump/Feedback
Direct Solar Heating/Feedback
Air Solar Assisted Heat Pump/Battery Storage
Solar Boosted Heat Pump/Feedback
Air Solar Boosted Heat Pump/Battery
COMBINED PV/THERMAL SYSTEMS
Rank
1
2
:3
4
IIIa,
IIIb,
IIId,
IIIc,
System Title
Solar Assisted Heat Pump/Feedback
Direct Solar Heating/Feedback
Stand Alone/Direct Heating/Battery
Solar Boosted Heat Pump/Feedback
Relative Effectiveness by Category (Collector) Type
Having ranked systems within a category, a comparative assessment was then made
of the relative effectiveness of collector types as represented by the three
categories. In this manner, similar systems appearing in more than one category
(i.e., having different collector types) could be compared prior to making
appropriate regional selections.
PV only solar arrays used in conjunction with an all electric load proved the
most suitable system for residential use. Their potential economic viability was
as good as, or better than, other solar energy options such as separate or
combined PV/Thermal systems in all regions studied. Feedback generally proved the
more cost effective approach when compared to battery storage for most systems
investigated. Battery storage achieved cost effectiveness in high insolation
areas, such as Phoenix.
It appeared that except at very low array costs, the higher thermal and
electrical efficiencies of the separate PV/Thermal panel systems more than com
... dnsate for the potential savings in structural and installation costs provided
by the combined PV/Thermal collectors. As a result, lower annual costs would be
2-5
possible with the separate PV/Thermal panel systems than with the combined
collector systems. Separate PV/Thermal systems also exhibited some advantages in
the Sunbelt areas, with the solar thermal systems capable of providing air condi
tioning, if appropriate solar cooling equipment is cost effective, or just
domestic hot water.
Combined PV/Thermal systems, while they may have roof space savings and may have
potential for lower cost in comparison to equivalent separate electrical and
thermal collectors, have exhibi ted poor performance in recent tests. Therefore
they currently do not offer significant promise.
The solar assisted heat pump (also referred to in the literature as a parallel HP
system) proved more cost effective and less complicated than a solar boosted heat
pump system (also referred to as a series HP system), with either the separate or
combined PV/Thermal collector.
System/Site Selections
The goal of the program was to have a set of site/system selections which covered
the major system options. Therefore, in addition to the results of the system
rankings and relative evaluations, a qualitative analysis was also used to narrow
the selected system configurations.
The array subsystem options were restricted to roof mounting due to the lack of
ground area for the arrays in a residential development. Only flat plate modules
were addressed. A roof-mounted concentrator was considered but eventually dropped
wi th Sandia concurrence. The type of flat plate roof mounting was considered a
key option since the different approaches required different installation details
and, therefore, varying installation costs. Thus, direct-mounted, stand-off and
integral array options received high consideration (See Figure 2-1) Rack-mounted
arrays were not included due to limited applications. Rack-mounted residential
arrays, however, were installed at the Southwest Residential Experiment Station
The reader is directed to that program for details of this mounting approach.
Similarly, the differences between feedback and battery systems received high
consideration and a separate design was developed for each. However, the dif
ferences in installation details between a stand-alone system and a grid
connected battery system were considered insignificant and separate designs were
nQt developed. PV-only systems with feedback were used in most of the remaining
2-6
designs since the systems have shown the highest potential in most previous
studies.
For PV/Thermal systems, side-by-side collectors were considered. A hot water only
system was addressed as an option of the solar assisted heat pump system
configuration. Solar cooling was eventually ruled out since solar thermal cooling
systems are not currently economically viable. Similarly, combined PV/Therma1
collectors were eliminated due to their current low performance and limited
region applicability.
For the power conditioning subsystem, installation details do not vary sig
nificantly for the different types of units available. Two primary options were
considered in the designs based on currently available hardware.
Finally, in addition to the three geographic regions initially considered, a
temperate climate with low space conditioning loads rounded out the weather
environment. House design variations appropriate for each of the environments
also were developed.
Summarizing all the considerations described, the six systems and regions identi
fied in Table 2-4 were selected. Subsystem options for several of the designs
were added to further expand the total coverage of hardware variations. Table
2-5 lists these options.
Figure 2-1. Alternate Mounting Techniques for PV Modules
2-7
Table 2-4. Selected System Designs
MountingSystem Configuration Configuration Location
PV-only with feedback Direct Southwest ,
Side-by-side PV/T Solar Assisted Heat Pump. Direct Northeast
PV-only with battery storage Standoff Southwest
PV-only for a passive house design Direct Northeast
PV-only with feedback Integral Southeast
PV-only with feedback Integral and Temperate ClimateStandoff
Table 2-5. Summary of Subsystem Options Addressed in the Designs
2-8
ARRAY SUBSYSTEM
GE PV Shingle Direct Mount Module(Two Design Generations)
ARCO Solar Batten Direct Mount Module
Solarex Stand-Off Mounted Frame Module
Side-by-Side Flat Plate Thermal Collectors
Integral Glass Laminated Modules
STORAGE SUBSYSTEM
Feedback Energy to Utility
Lead Acid Battery Storage
HVAC SUBSYSTEM
Heat Pump (HP)
Fossil Heating/Vapor Compression Cooling
Solar Assisted HP
Solar Domestic Hot Water (DHW)
Space Conditioning Heat Recovery for DHW
POWER PROCESSING AND CONTROL SUBSYSTEM
Windworks Gemini Inverter
Abacus Sunverter
DESIGN NUMBER
1 and 4
2
3
2
5 and 6
1, 2, 4, 5 and 6
3
1, 2, 3, 4, 5
2
2
2
3
1, 3 and 4
2, 5 and 6
SECTION 3
SYSTEM DESIGN SUMMARIES
A new house design was developed for each application. The designs were all
single familiy detached houses but they included one-story, two-story and zero
lot-line type configurations. The floor areas ranged from 142 m2 to 161 m2
wi th south facing roof areas ranging from 44 m2 to 104 m2 • All the houses
were assumed newly constructed in 1986. The designs included basic energy
conservation features and additional passive design features projected for 1986
but consistent with the aesthetic design of the house.
The electrical energy derived from the PV system serves the normal household
electrical requirements including general appliances, lighting, cooking and hot
water heating. In general, the homeowner was considered the system user; there
fore, operation and maintenance requirements typical of conventional HVAC systems
were assumed. The systems excluded instrumentation, since a mature system in
stallation was assumed. Specifications for equipment were based on currently
available components or similar currently available equipment if the component
was not available.
To evaluate the performance of the system designs, typical hourly electrical load
profiles and space condi Honing demands for the house were used. These load
profiles were developed in the GE Regional Residential Study, Reference 5, and
updated in the Residential Load Center Program, Reference 14. All the system per
formance analyses were completed based on these hourly loads and using Typical
Meteorological Year (TMY) weather data as developed by Sandia National Laborato
ries. Life cycle cost analyses, based on system performance results and system
cost estimates, were used for system sizing and tradeoffs.
The six designs developed included:
1. A PV system for an all-electric Southwest residence employing a direct mountarray
2. A PV/thermal side-by-side system for a Northeast residence
3. A PV system with battery storage for the Southwest employing a standoff array
4. A PV system for a passive house design for the Northeast
5. A PV system for a Southeast residence employing an integral mount array
6. A PV System for a temperate climate with either a standoff or an integralmounted array
3-1
• System Life: 20 years
• Maintenance: $lOO/Year
• 1980$
• Battery Life: 10 Years
A separate report was written for each design, References 3 through 8. A complete
set of full sized drawings were also delivered to Sandia for each of the designs.
The system sizes ranged from 4 kW to 8 kW systems. The system sizing for all the
designs was completed on a marginal cost basis. A consistent set of economic
assumptions was used and they are summarized in Table 3-1.
All of the economic calculations were completed for a 1986 start in constant
"1980$. The overall average inflation rate of 5 percent was assumed through the
time frame of the analysis. This value is low according to cur.rent. rates but
since the marginal cost analysis becomes independent of the inflation rate and to
maintain similarity to previous work, this value was used. In addition, a
resul tant homeowner mortgage rate of 10 percent for a 20-year loan was assumed
with a marginal income tax rate of 35 percent for the homeowner. This is
equivalent to a taxable income of approximately $25,000 at present tax rates.
Since several states and local communities have already exempted solar systems
from property tax, no additional property tax were assumed. These figures imply
an annual cost or "fixed charge" of about 9.1% of the initial photovoltaic system
cost. Annual operating costs were assumed to be $lOO/year for operation and
maintenance and 0.5% of system cost for insurance. All components of the solar
system were assumed to have the same lifetime as the system (20 years) except for
batteries (10 years).
An electricity price escalation of 4% over inflation was also used in the
analysis, although the marginal benefit/cost analysis allows the extension to
other escalation rates and system lifetime assumptions by a constant adjustment
factor as done in Reference 16.
Table 3-1. Economic Assumptions for Optimization
• 1986
• General Inflation Rate: 5%
• Insurance: 0.5% of Capital Cost
• Electricity Price Escalation: 4% Over Inflation
• Mortgage Rate: 10%
• Tax Bracket: 35%
• No Property Tax
3-2
.'
Cost estimates were made for each. of the systems to calculate the levelized
annual costs. The array cost assumed was the National PV Program goal of 70i/peak
Watt or $700/kWp factory price. Including the markups, the cost is $925/kWp on
site. The balance of system costs were made up of array installation estimates,
power conditioning subsystem costs, storage subsystem costs and remaining
equipment costs. Power conditioning costs were based on 1986 price projections.
The 1986 battery costs were varied from $O/kWh to $150/kWh to assess the effect
of these values on system sizing. All the costs are in 1980$ and include in
general, two 15% markups for distribution and contractors. These markups are
probably low, but were used in the analysis until more detailed numbers are
available for distribution networks.
Regional energy price estimates were assumed for each design to determine the
system-levelized annual benefits.
The following subsections provide a brief summary description of each of the
designs.
3-3
A PV SYSTEM FOR AN ALL-ELECTRIC RESIDENCE IN THE SOUTHWEST
A simple, utility feedback, photovoltaic system was designed for a single story
all electric house in the Southwest. The two major system elements are the photo
voltaic array and the power conversion subsystem. The photovoltaic array uses the
GE Block IV shingle module in a direct mount configuration. The system was sized
nominally at 8 kWp based on minimum life cycle costs but key parameters were
evaluated parametrically and changes in some of the assumptions were also
evalua=ted. For example, sellback rates were varied parametrically indicating at
higher sellback rates (greater than 50%) all of the available roof area should be
utilized for the array, resulting in the 8 kW sized system. As the sellback rates
decreased to 30% or less, a 5 kW system size becomes more appropriate. Assump
tions of fixed and variable system capital costs also affect system sizing,
implying smaller system sizes as fixed costs are lowered.
The system design is applicable to all regions. The effects of varying electrical
load level on the system economics resulted in cost-to-benefit ratio insensitivi
ty to load level at sellback rates above 50%. Therefore, system size would not be
affected by slight variatione in load levels. Profile changes and dramatic
changes in the load would affect system sizing.
3-4
HOUSE DESCRIPTION
• The house design is for a SINGLE-STORY residence of NEW CONSTRUCTION forthe Southwest region of the country.
• The design includes PASSIVE SOLAR FEATURES and ENERGY CONSERVATIONFEATURES projected for 1986.
• There is 149 m2 (1600 ft2 ) living area with 104 m2 (1120 ft2 )south facing roof area.
• The house is ALL ELECTRIC with a 3-ton heat pump and electric hot waterheater.
• The site layout has a detached garage with a lot area between 1/6 and1/4 acre.
3-5
SYSTEM DESCRIPTION
• The system is grid connected with an 8 kW NOCT array rating using a GEBLOCK IV SHINGLE MODULE ARRAY. The array consists of a total of 475MODULES COVERING 93 m2 in a redundant parallel-series network.
• The power conversion subsystem uses a 10 kVA LINE COMMUTATED MAX POWERTRACKING INVERTER to convert dc genera ted power to ac. A 15 kVA SINGLEPHASE ISOLATION TRANSFORMER is used to match ac supply voltage to theload.
• The system operation is PARALLEL AND SYNCHRONIZED WITH THE UTILITY.
• Excess generated power is FEEDBACK to the utility.
• The system represents the SIMPLIEST PHOTOVOLTAIC DESIGN with a minimumof components and controls.
PVARRAY
UTILITY BACKUP
3-6
- ROOF MOUNTED SHINGLE93m2
Dc/ACINVERTER
MAX POWERTRACK
GENERALLOAD$
HEATPUMP
HOTWATER
SYSTEM OPERATION
• The system has automatic startup and shutdown control.
• The system automatically shuts down with loss of the utility.
• System operation is summarized by the sequence below.
1. At sunrise in the automatic "on" mode, the ac and dc contactorswill close when the array bus voltage reaches a threshold of 180Vdc.
2. During the daylight period the inverter will continue to operate aslong as there is a net power output.
3. The inverter will track the maximum power operating point within +1percent over the range of 180 to 220 Vdc.
4. The interruption of utility-supplied power will eause the dccontactor to open and remain open until line voltage is restored.
5. At sunset, the inverter ae and de contactors will open when the netpower output falls to zero. These contactors will remain openthroughout the night to eliminate the majority of the inverterparasitic losses.
ununRIlYICII-3 WIRE, SlNGLI _ .._VA(;
AOOFAR.RAY2!5S a 19PSHINGLE
POWER COOIVERIION SYSTEM St~VIC£ PI\lI£l
3-7
PHOTOVOLTAIC ARRAY
The array consists of shingle PV modules connected in a 25 series by 19parallel network covering 93 m2 of roof area.
• The array is oriented due south with a roof. pi tch of 260. The overallcell packing efficiency is 76.3% over the 93 m2 •
• The modules are direct mounted on top of the roofing felt and plywoodroof sheathing. They form a weather tight roof.
• The modules are installed by an overlapping procedure similar to conventional shingles. Four electrical interconnections are made with flathead machine screws per module and two roofing nails are used per modulefor attachment to the roof.
STARTER COURSE
3-8
ROOF EDGE
FINISHING ROW OFDUMr1Y EDGESHINGLES
POSITIVE TERMINATION SHINGLES
PHOTOVOLTAIC MODULES
• The module was developed by General Electric Company as part of the JPLBlock IV procurement.
• The module uses 19 ARCO-SOLAR 100 mm cells with an unencapsulated efficiency of 12.3% connected in a series circuit.
• For a NOCT of 640C, the maximum power output is 17.14 Watts and 7.3Volts at SOC conditions (1 kW/m2 , 200C ambient, 1 m/s wind speed).
• A summary of module characteristics include:
Module weight: 3.85 kg
Total cell area: 0.1492 m2
Exposed module area: 0.1955 m2
Module packing factor: 0.763
3-9
POWER CONVERSION SUBSYSTEM
• The PCS provides the interface between the PV array and the normalresidential utility service and loads.
• The subsystem consists of three main components: the inverter, the dcfilter and transformer along with the associated control circuitry.
• The subsystem is sized for 10 kW of power output with a 15 kVA transformer sized to accommodate the out of phase ac voltage (VARS) andcurrent.
• The subsystem can be ottained from the Gemini Corporation, marketed byWindworks, Inc.
KEY INVERTER DESIGN CHARACTERSTICS
OUTPUT POWER RATING
OUTPUT VOLTAGE:
INPUT VOLTAGE:
FULL LOAD POWER FACTOR:
FULL LOAD EFFICIENCY:
FULL- LOAD HARMONIC DISTORTION
OPERATING TEMPERATURE
10 kW CONTINUOUS
240 VAC Utility Residential Service
200 + 20 Vdc
60% Minimum,
92% Minimum
30% Maximum
00 to 400 C
POWER CONVERSION SYSTEM
Voc--~
'oc---.l"---_.......
3-10
PV INTERFACE WITH UTILITY AND HOUSE SERVICE
• Interface arrangements employ CONVENTIONAL WIRING RUNS AND EQUIPMENT asmuch as possible to facilitate acceptance by local regulatoryauthorities.
The PV array source is treated as a CONVENTIONAL UTILITY SERVICEentrance to the residence with the raceway parallel to utility line.
• An external disconnect switch provides an external break. By strict codeinterpretation, this ~witch may be eliminated, especially as PV installations increase.
• An equipment room of 6.1 m2 floor area is located on the west end ofthe house. The equipment room will also house the heat pUmp and electrichot water heater and can be used for extra storage.
• All PV related equipment is wall mounted, except for the transformer.
POSITIVE BUS BARPENETRATION
~-----STANDARO WEATHERHEAD~-----~IN LOOP
l"&' .('IIII STANDARD ENTRANCEr--1 FOR UTI LI TY: SERVICEIIII1II
[~[/I DISCOIINECT .,roJlSWITCH~~i~==~~
SOLAR COLLECTING ARRAYdCONDUIT RUN
I
I--------------------------~
3-11
ARRAY SIZING
• Energy sellback price the utility is willing to pay the homeowneraffects system sizing.
• Sellback rates greater than 50% imply full roof arrays (93 m2 ).
• Lower sellback rates (-30%) imply roof arrays of 50 m2 •
• At higher sellback rates, the economics are insensitive to load level.
ALBUQUERQUE PHOENIX
oo 20 40 60 SO 1 0
COLLECTOR AREA, M2
1.6
1. II0 SELLa- BACK< 1.2 lli!.Q PS/PE~
t: 1.0u. Iw2 I 0.3wen 0.50I- 0.1l- IIII0 ..... OESIGNU
I POINT
I
r-0ESIGNPOINT
IIII
O.SI
0.7
SELLBACK~ PS/PE1.6
oI<ex: 1.2I-
ffizwenol-IIIIoU
PHOENIX
SELLBACK
_____ RATIO PSI?E
=====-;-;;;;;;~::0.30.5
0.7
1.0
1.8
1.6
2I- 1.11<Ill:
!: 1.2Il.
'" 1.0z'"(ll0 0.8l-I-III 0.60U
O. II
0.2
0
0
3-12
2Array Area = 92.9 m
'0.5 1.0ACTUAL LOAD/NOMINAl. l.OAD
1.5
DESIGN PERFORMANCE
• Net system output for both Albuquerque and Phoenix is greater than thetotal electrical load.
• Phoenix shows better load matching characteristics.
• Overall· system efficiency based on incident insolation for gross arrayarea is within the 8% to 8.4% range for the Southwest.
PHOENIX ALBUQUERQUE
UTILITY MAKEUP ·9 MWHUTILITY FEEDBACK -12.6 MWH%OF LOAD SUPPLIED
DIRECTLY -39%
PV LOADSYSTEM 14.8 MWHOUTPUT18.3 MWH
J FMAMJJASOND
1.5
0.5
2.0
::r:::~
>'oa:~ 1.0w
NET FROMUTILITY,......------------.
UTILITY MAKEUP· 8.7 MWHUTILITY FEEDBACK -10.3 MWH%OF LOAD SUPPLIED
DIRECTLY -45%
JFMAMJJASOND
2.0
1.5
o
0.5
1.0
::r:::~
>'oa:wzw
3-13
SIDE-BY-SIDE PV/THERMAL SYSTEM FOR THE NORTHEAST
A side-by-side photovoltaic/thermal system with utility feedback was designed for
a two-story all-electric house in the Northeast. The three major system elements
are the photovol taic array, the electric power conversion subsystem, and the
thermal subsystem. A key to the sizing of the PV and thermal array is the
relative cost of each system. Therefore, the ratio of the costs associated with
each system was varied parametrically to define ultimate sizing of the PV and
thermal array. In general, the results indicate larger PV array area and minimal
thermal array area for the set of cost assumptions. A thermal system of adequate
size, however, was selected to assure sufficient design details for determining
system installation costs. The PV system was sized nominally at 6.7 kWp based on
roof area limitations, side-by-side with a 19.5 m2 flat plate thermal collector
array. The 73.6 m2 photovoltaic solar array uses a batten type solar cell
module, being developed by AReO-Solar as part of: the JPL block IV procurement,
and the 19.5 m2 thermal array uses a Sunworks SOlector@ collector which is
representative of available thermal collectors.
A separate thermal system design was also ~eveloped for a side-by-side solar hot
water system for comparison. A solar hot water system, since it was less complex
and has less components, showed better economic viability than the parallel solar
heat pump space heating and hot water system. Both systems probably will be
available in the 1986 time frame.
3-14
HOUSE DESCRIPTION
• The house design is for a TWO-STORY residence of NEW CONSTRUCTION forthe Northeast region of the country.
• The design includes PASSIVE SOLAR FEATURES and ENERGY CONSERVATIONFEATURES projected for 1986. A GREENHOUSE is attached on the southernexposure.
• There is 158 m2 (1700 ft 2 ) of living area with 101 m2 (1090 ft2 )south facing roof area.
• The house is ALL ELECTRIC with a 3-ton heat pump and electric hot waterheater.
• The site layout has a two car attached garage with a lot area of approximately 1/4 acre.
3-15
SYSTEM DESCRIPTION
• The system consists of a separate GRID CONNECTED photovoltaic array witha 6.7 kW NOCT rating and a thermal array which provides a SOLAR ASSISTto the HEAT PUMP and DOMESTIC HOT WATER HEATER.
• The PV array uses the ARCO SOLAR BLOCK IV BATTEN SEAM MODULE in a 28series by a 4 parallel circuit arrangement covering 73.6 m2 of roofarea.
• The thermal array uses19.5 m2 of roof area.
10 Sunworks@
Solector collectors covering
• The power conversion subsystem uses a 8 kW SELF-COMMUTATED MAX POWERTRACK-ABACUS INVERTER to convert dc generated power to ac and to matchac supply voltage to the load.
• The system operation is PARALLEL AND SYNCHRONIZED WITH THE UTILITY.
• Excess generated power is FEDBACK to the utility.
• The thermal subsystem has a 350 GALLON WATER THERMAL STORAGE componentand an 80 GALLON DOMESTIC HOT WATER TANK.
BACK·UP
PV Dc/AC I- GENERALARRAY INVERTER
~
LOADS
MAX. POWERTRACKER f- HEAT
~r4-PUMP
- HOT -~WATERTHERMAL ~ THERMAL
ARRAY. $Tt)~~t;F.
3-16
PV SYSTEM OPERATION
• The system has AUTOMATIC STARTUP and SHUTDOWN control.
• SYSTEM OPERATION is summarized by the sequence below.
At sunriseclosed andenergized.
in the automatic "on"the phase lock loop
mode, the ac line contactor isand maximum power tracker are
2. During the daylight period, the inverter will continue to operateuntil the dc input voltage falls below 160 V.
3. The inverter will track maximum power point wi thin +1 percent overthe range of 160 to 240 Vdc.
4. The interruption of utility-supplied power will cause the accontactor to open and remain open until line voltage is restored.
5. At sunset, the inverter ac contactor will open when the net poweroutput falls to zero. This contactor will remain open throughout thenight to eliminate any parasitic losses.
UTILITYSERVICE
INVERTER
EXTERIORSWITCH
INTERIORSWITCH
SWITCh
MAINCS
VARiSTOR
PVCB
TYPICALSRANCH
SERVICEPANEL
RESIDENTIALLOAD
3-17
THERMAL SUBSYSTEM
.. The system has AUTOMATIC STARTUP AND SHUTDOWN control.
A control centerenergy collection,backup.
regulates operational sequences for solar thermalstorage and utilization loops, as well as auxiliary
o SYSTEM OPERATION is summarized by the sequence below:
1. The collector and storage loop pumps are activated when a temperature differences of S.30 C is registered between the collectorsurface and the bottom of the thermal energy storage (TES) tank.
2. Operation continues until the temperature difference is less than2 .SoC.
3. Flow is diverted through the heat pump when the TES temperature isgreater than S7.SoC
4. Circulation is activated from the TES to the domestic hot water(DHW) tank when a temperature difference of 3.30 C exists betweenthe top of the TES tank and the bottom of the DHW tank and continuesuntil the temperature difference falls to less than 1.loC.
5. Flow is activated to the hydronic coil when the TES temperature is26.7°C or greater. This is an adjustable setting.
6. If the hydronic coil cannot meet the space demand, then the flow tothe hydronic coil is stopped and the heat pump activated andoperated in a conventional mode.
EXPANSION 12RTANK
HYDRONICCOIL
SOLARCOLLECTORS STORAG
HEAT HEAT TANK
DUMP EXCHANGER (TES) (P3) HP gPUMPCOIL
EXPANSION TANK HOT WATER c:><===> FAN
TANKTSUP '0PUMP (P2) /
/......./.......
"- /.... - - TGR
DIFFERENTIALCONTROLLER
3-18
PHOTOVOLTAIC ARRAY
• The array consists of batten PV modules connected in a 28 series by 4parallel network covering 73.6 m2 of roof area.
• The array is oriented due south with a roof pitch of 400 •
• The modules are direct mounted on top of the roofing felt and plywoodroof sheating. They form a weather tight seal.
• The modules are installed by an ovelapping procedure similarconventional batten seam roof. The modules are held in place byscrewed into the roof. Electrical interconnections are madepatented connectors which run along the batten seam.
to aclipswith
...~ ... - C"~"'I'.& ,.pv.::. - __
L:I .. 1..1 l.- i..
)-(.-- - ....-- r--- - - - . r-
++ ++ ++ ++ ++ ++ ++ ++ ++ ++
'-'r ~-TWl'-u--........J..
~,~e..-La- ,.."..."CO •
=~~~+ +~ + +~
.;:.:,toc.=
.-- -++ + +-1-.--++~ ++
-- .. --10.
'='::io'I'=-='JL.::..::1L:..::J- ..- -11..++ + + + +Iol + +
-++
- - ..~J
L::,,::-=-.:-=.nll c-~-~Io. -=-=....-- -++ ++
~ -.-ItI --1'1..::...::,..-- _1 __++Iol ++Iol ++Iol + +•
c..::~'='::Jc..::J~r- ,..-- .-- r-++ ++ ++ ++
~1o..=.=1o...::.::.1ol=-= =-= ~"'=-="c..:.J-...:.:...r___.---- __++~ ++ ++ItJ++al ++ ... ++1oסi ++ ++
~~+ ...U
~ .::..::~ ~ L::..::~'::":J.::..:J..:.:J-- ..:.: -=-=: I ::.:J-=-= ~ -=-= - -J-=-=J -=-=J- - .-- .-- - .-- r-- - .-- __ .-- r- 1.-- ...-- .--f++'" ++ ++~ ++100 ++ ++ ++ ++ ++ ++ ++ ++~ ++IJ ++ ++ ++
3-19
PHOTOVOLTAIC MODULES
The module was developed by ARCO Solar, Inc. of Chatsworth, California,as part of the JPL Block IV procurement. The development modulecurrently uses circular cells.
The module specified for the design is a high density version which usesARCO-Solar 70 mm square cells with 16 cells connected in series and 7parallel circuits for a total of 112 cells per module.
For a NOCT of 630 C, the maximum power output is 60.0 Watts, and 6.0Volts at SOC conditions (1 kW/m2 , 200 C ambient, 1 m/s wind speed).
A summary of module characteristics include:
Module weight: 4.8 kg
Total cell area: 0.5488 m2
Exposed module area: 0.657 m2
Module packing factor: 0.835
Nominal size: 1.2 m long by0.58 m wide
3-20
THERMAL COLLECTORS
• The thermal collector is a flat plate model called the Solector andit is currently available from Sunworks of New Haven, Connecticut.
• The collector uses a SELECTIVELY COATED COPPER absorber, with a SINGLEGLAZING of low iron, 1/4 inch' tempered glass in an extruded ALUMINUMframe.
• The collector has INTEGRAL MANIFOLDING for ease of installation.
• The design installation consists of 10 collectors covering 19.5 m2 •
• A summary of collector characteristics include:
Collector size: 213 m long by 0~91 m wide
Collector weight: 51.4 kg
Effective absorber area per panel: 1.7 m2
.Liquid-Cooled Solector~, Black Chrome Coating
tOO
.10
.80r
~ 10z..~ .eo...... .50
~! 040
~~ .30
! .20
~ f-SINGL GLAZED·- ~
~
~DOUBLE GLAZED - ~
.10
00o .05 .10 ~ ~ n ~ • ~
FLUID PARAMETER (Tj-T.I/ql.OF/BTU/1l2 .,
,45 .50
Single glazed, lo·iron: Double glazed, lo·iron:
Av. flow rate, 278.2Ibs/hr.; Av. mass flow 278.85Ibs/hr.;avo ambo temp. 73.60 of. avo ambo temp. 100.56°F.(x) Field data (0) Field data
Tests conducted by Soiar Energy Analysis Laboratory; Desert Sunshine Exposure Test,lnc.
Data from Sunworks Brochure
3-21
POWER CONVERSION SUBSYSTEM
• The Power Conversion Subsystem (PCS) provides the interface between thePV array and the normal residential utility service and loads.
The subsystem consists of three main components: the inverter, the dcfilter and the transformer along with the associated control circuitry,packaged in a single unit.
• The uni t is manufactured by Abacus Controls, Inc., Sommerville, NewJersey.
• The subsystem is sized for a 8 kW POWER OUTPUT with the transformersized to accommodate the adjustment of the ac voltage and current.
Characteristics are summar~zed below.
KEY INVERTER DESIGN CHARACTERISTICS
OUTPUT POWER RATING:
OUTPUT VOLTAGE:
INPUT VOLTAGE:
FULL LOAD POWER FACTOR:
FULL LOAD EFFICIENCY:
FULL LOAD HARMONIC DISTORTION:
8 kW CONTINUOUS
240 Vac Utility Residential Service
200 ~ 40 Vdc
Unity
90% Minimum (over 25% to 88% ofoperating range)
5% Maximum
MAXIMUM (+)y~(_) INPOWER 1---_.-1~ ~--t PHASETRACKER SENSE
AMPLITUDECONTROL
,AC LINE
ROOF DC CONTACTORARRAY FILTER I SCR I XFMR nr'll"\ n n :1- AC OUTPUTINPUT I TNV :~TI=~ I vv~
I - SENSE
9 'AN( THERMAL
PLL SWITCH UTILITYV - REFERENCE VOLTAGE
3-22
PV INTERFACE WITH UTILITY AND HOUSE SERVICE
• Interface arrangements employ CONVENTIONAL WIRING RUNS AND EQUIPMENT asmuch as possible to facilitate acceptance by local regulatorYauthori ties.
• The PV array output is treated as a CONVENTIONAL UTILITY SERVICEENTRANCE to the residence with the raceway parallel to utility line.
• An external disconnect switch provides a means to externally disconnectthe array and the power conversion unit. By strict code interpretation,this switch may be eliminated3 especially as PV installations increase.
• An equipment room of 8.2 m2 floor area is located on the west end ofthe house. The equipment room also houses the heat pump, electric hotwater heater and thermal storage tank and can be used for extra storage.
• The Power Conversion Unit is floor-mounted, and all of the remaining PVrelated equipment is wall-mounted •
Array
,.,-----tMa inta in 12 11 Cl earanceBetween Services______~1#4 AWQ. Positive-Conductor
Pl us #4 'AHc;i. Negat i veConductor Run in 111 PVCConduit
~~~+-~~~ Standard Service -Entrance Conductorsfor Ut il ity
.Thermal Collectors
Weatherhead for Utility~~~-----I'Service
~ . Strap Supports in·T.a~ Accordance with Disconnect
L... N.E.C. (Typical) ---'Switch -------- •. --
P,V, SystemConnect ion from ---:~~H+-'!"""""H+THW,
Attic
3-23
ARRAY SIZING
• Energy sellback price the utility is willing to pay the homeowner andrelati ve PV system cost to thermal system cost affects the overallsystem sizing.
• Sellback rates greater than 50% imply full roof arrays.
• Marginal costs were used in the system sizing.
• For thermal system costs greater than PV system costs, system sizingresults in more PV array area and less thermal collector area.
• Thermal collector area was sized primarily to provide a thermal systemof adequate size for cost estimates.
o
MADISON
'1.5
1.4
~1.3
1.2Cll:uu-' 1.1
PV ARRAY AREAI
1.0 73.6~52.6m2~
I PV ARRAY AREAI
73.6nL.,...-s2.6m2-h,.5mZ-,j1
II
I
1.5
1.2
1.1
1.4
10 20 30 40 50 60
THERMAL COLLECTOR AREA, m210 20 30 40 50 60
THERMAL COLLECTOR AREA, m2
PV System Cost "Fpy +CpV x Areapy
py System Cost -2$1603 + $131/m x Area pv
Thermal System Cost"FT + CT x AreaT
Thermal System Cost.C
$4612 x cT x S131/m2ArPV
Base CaseCTc" 1.7P
Cr/Cpv " 2
CT/CpV • 1
CT/Cpy • 1121
1- DESlr,rl POI~TII
20 30 40 50 60GROSS THERl1AL COLLECTOR AREA. m2
10o
1.0
1.8 Ps/PE" •5
1.6
1.4Cll:uu-' 1.2
o 10 20 30 40 50 60
: ROOF AREA (BASED ON 95 m2)
3-24
DESIGN ELECTRICAL PERFORMANCE
• Net PV system output for both Boston and Madison is approximately 67% ofthe load requirements.
• Approximately 39% of the PV output is utilized directly in the house andthe remainder fed back to the utility.
• The net electrical system output in the summer months is greater thanthe load requirements, mainly due to the thermal system providing thehot water requirements.
• Overall PV system efficiency based on incident insolation for grossarray area is approximately 8.8% for this design in the Northeast.
0 0J F MA MJ J A S 0 N D J F MA MJ J A SON D
ANNUAL TOTALS ANNUAL TOTALS
• LOAD J2~7 M\~H • LOAD 14.~ MWH
• PV SYSTEM OUTPUT §'R M\'IH • PV SYSTEM OUTPUTl~:~
MWH
• UTILITY MAKE-UP MWH • UTILITY MAKE-UP MWH
• SELL BACK ~:2 MWH • SELL BACK MWH
3-25
DESIGN THERMAL PERFORMANCE
• Net thermal system output ranges from 35 to 39% of the load requirements.
• Approximately 24% of the space heating requirements and 66% of the DHWrequirements are supplied by the solar thermal system.
• All of the thermal requirements are met in the summer months.
.. Overall thermal system efficiency is approximately 22% for this designin the Northeast.
3.0
::c: BQS.IOli MADISON::c 2.'- 2.5~
• SPACE ::c: SPACE>- HEATING ~ 2.0 HEATING~
Ic::uJ >-z 1.5LI.! U)
0:;.
>- uJ~ ziE w
1.0z >-0 ~
~ ::c:I- 0.50.5 z0~
0 0J F M A M J J A S 0 N D J F M A M J J A S o N D
~ 0.5 ~~::~c HOT:::::0[; "d '.I~
J F M A M J J A SON DANNUAL TOTALS
• SPACE HEATING LOAD 9,76 MWH• SOLAR TO SPACE HEATING 2444 MWH• DOMESTIC HOT WAIER LOAD .75 MWH• SOLAR TO DHW 3, / MWH
. LOADSOLAR SUPPLYo'Sh; DOMESTIC H~O.T~...TER
~ -_ ..~~--::: _- _-~ -.L!--I.'_---! I-..I.' -l
J F MA M J J A SON D
ANNUAL TOTALS
• SPACE HEATING LOAD 13.07 MWH• SOLAR TO SPACE HEATING 2.96 MWH• DOMESTIC HOT WAT~R LOAD 4.89 MWH• SOLAR TO KHW 3.22 MWH
3-26
A PV SYSTEM WITH BATTERY STORAGE FOR THE SOUTHWEST
A photovoltaic system with on-site storage was designed for a single story all
electric house in the Southwest. The three major system elements are the photo
voltaic array, the battery storage subsystem, and the power conversion subsystem.
Marginal life cycle cost analysis was used for system sizing of the array and
battery. The analysis considered various combinations of array and battery sizes
and provided parametric performance results. In general, the results indicate a
relatively small battery capacity for each array size considered, based on
current cost estimates.
The nominal size array output is 6.07 kWp. The modules are standoff mounted. The
baseline design battery capacity for Phoenix is 20 kWh. An Albuquerque applica
tion, where increased electricity cost and improved load match exist, requires a
larger capacity of 25 kWh. Reduced battery capttal costs indicate an increased
optimum capacity for all of the array sizes considered. The array size and
associated array performance ultimately bound the battery capacity, if cost con
straints are neglected. The system economics also indicate a sensitivity to total
electrical load requirements for all battery costs considered. Therefore, changes
in the load requirements and profile affect system sizing and particularly the
optimum battery capacity.
3-27
HOUSE DESCRIPTION
• The house design is for a SINGLE-STORY residence of NEW CONSTRUCTION forthe Southwest region of the country.
• The design includes PASSIVE SOLAR FEATURES and ENERGY CONSERVATIONFEATURES projected for 1986.
• There is 149 m2 (1600 ft2 ) living area with 104 m2 (1120 ft2)south facing roof area.
• The house is ALL ELECTRIC with a 3-ton heat pump and electric hot waterheater.
• The site layout has a detached garage with a lot area between 1/6 and1/4 acre.
3-28
SYSTEM DESCRIPTION
• The system is grid connected with a 6.07 kW array rating using a SOLAREXBLOCK IV INTERMEDIATE LOAD MODULE ARRAY. The array consists of a totalof 100 MODULES with 76.2 m2 of aperture area in a parallel-seriesnetwork.
• The battery storage subsystem includes a 20 kWh LEAD ACID BATTERY tostore PV generated power. A BATTERY CHARGE CONTROLLER controls the busvoltage. A larger capacity of 25 kWh is required for an Albuquerquelocation.
• The power conversion .subsystem (PCS) employs a 6 kVA LINE COMMUTATEDINVERTER to convert PV generated power to ac. A 10 kVA SINGLE PHASEISOLATION TRANSFORMER matches ac supply voltage to the load.
• The system operation is PARALLEL AND SYNCHRONIZED WITH THE UTILITY,without feedback.
• Excess generated power is SHUNTED to ground. Heating domestic hot waterwith the excess energy is a design option.
UTI LITY BACKUP,PV DC/AC GENERAL
ARRAY INVERTER LOADS
--SHUNT - HEAT-
IPUMP
1- - HOT- '----WATER
3-29
SYSTEM OPERATION
• The system has automatic startup and shutdown control.
• The system automatically shuts down with loss of utility power.
• System operation is summarized by the sequence below.
1. At sunrise, in the automatic "on" mode, the ac and dc contactorswill close when the array bus voltage reaches a threshold of 120Vdc.
2. During daylight hours the PV array/battery system will supply theload demand. Excess power is applied to the charge battery.
3. When the battery is fully charged (156 Vdc), the battery chargecontrol sheds PV branch circui t( s) which reduces current flow~
Should the voltage fall below 144 Vdc, the charge control addscireui t( s) •
4. The battery delivers load demands when the voltage level is greaterthan 120 Vdc and the solar array output is not adequate to meetthese demands.
5. The battery supplies power until the bus voltage falls below 120Vdc; at this point the ac and dc contactors will open.
6. If the battery is depleted, a 128-minute time lag is imposed beforeanother discharge attempt is made, thus precluding deep discharges.
7. Twice a month, during off-peak hours, the pes initiates batteryequalization to 173 Vdc.
ISOLATIONDIODES
.....-- FILTER POWER CONVERSION SUB
..!',l.FUSED
OUTDOOR INDOOR I
IAC.... BREAKER DISCONNECT TRANSFORMER I PANELBOARD
SOLAR • r7l r:;-ARRAY • I __ I I
-+ '/)A
~llf• -1>1- I L N
FUSED I,J. ,! I INVERTER'/)B I
'1:7 DISCONNECT
IAND ----l )'--- BATTERY CHARGER,
I••• ANDTIMING AND I
DI CONTROLS
-~~
BATTERY I II~ I LIGHTNING"\-- PROTECTOR
SHUNTS CHARGE
\.CONTROL
BATTERY ~
\INTERLOCK
VENTILATIONSYSTEM
\
3-30
PHOTOVOLTAIC ARRAY
• The array consists of STAND-OFF PV modules connected in a 10 SERIES by10 PARALLEL NETWORK with 76.2 m2 of aperture area.
• The array orientation is due south with a ROOF PITCH OF 260 • Theoverall module packing efficiency is 85.3% over the 76.2 m2 •
• The modules are mounted on WOOD 2 x 4 inch STAND-OFFS. Conventionalasphalt shingles beneath the modules provide a weathertight roof.
• The module frame fits into aluminum clips mounted to the stand-offs andattaches to the clips by four sheet metal screws. Electrical interconnection between modules is made with AMP Inc. connectors and standardcables.
10
UOTE::-:,:
1. ?P~'-IVc:- ~;::>uc..T~ ;;eo....... c.Ao:::.i-leul...l --:'""c:;:o' I~Qt...~TI e>o-i :::::>.or:>E:- e:".::;o"""",'v,~ l--l1e:...b~A."""
GENERAL" ELECTRIC...... DIYlelON
Detaillld RetidInlIII P.Y. system
~~~ooc.#13-8779
'Johnson 8& StoIIer,lnc.8ectrica1 ~ineera127 Tauntonst.MiddlebottJugh,Mase. 02346617-947-8464
. .masS\...~slgn--"-1:II"_""e-.-._o:tI.,..,__
SouthwestAll-Electric HouseWith P. V./Battery
..... Ho~
-....
3-31
PHOTOVOLTAIC MODULES
• The module is manufactured by SOLAREX CORP., of ROCKVILLE, MARYLAND, andwas developed as part of the 3PL Block IV procurement.
• The module uses Solarex 95 mm SQUARE CELLS with 36 cells connected inseries and 2 paralle~ circuits for a total of 72 cells per module.
• For an estimated stand-off NOCT of 60°C, the MAXIMUM POWER OUTPUT is60.7 Watts, and 13.9 Volts at SOC conditions (1 kW/m2 , 20°Cambient, 1 m/s wind speed).
• A summary of module characteristics include:
- Module weight: 15.9 kg- Total-cell area: 0.650 m2- Exposed ;module area: 0.762 m2
- Module packing factor: 0.853- Nominal size: 1.2 m long by 0.64 m
wide1--------------116.8 CM------------..1/
I
~I 'I il ,i Ir--l II :1 \\ I
, !i 1 I, :: il : j 11 ~ I
iLl :, I i! L! ~ lLJ-.JIi II i,l 1\1 il iI' II' lUi1:1' II il ii',
60.4 eM; ,! :, I !I ~, I I 'I". ...--,-t---t-------l- --- ---; I'U,'~ II iIi IUil 'III II I! ill' 1'1'
• I II :I L I 1 , , 1\ I ..
I ~ [-1 ,II,: liill~,ri,nl-, '!L-!U 'L...J! :1 I, :1 Ii ;1 i
'\ \,1 I' l' ':--11 II -:r-1,1",i!I---;I-'ij1I I 'I I '! II ' II I, ;1 ! ,
II II !! II I \I!I II Ii II ILirop VIE:Wr.!------ '120 Q; ---------------------"1
3-32
63.5 eM
I
,'h-../1 l=--~II.
TerminalConnectors
o 9
:1I·j
POWER CONVERSION SUBSYSTEM
• The Power Conversion Subsystem (PCS) provides the interface between thePV arr~/batter,y and the normal residential utility service and loads.
• The subsystem consists of three main components: the INVERTER, the DCFILTER and TRANSFORMER along with the associated control circuitr,y.
• The subsystem rating is 6 kVA of power output with a 10 kVA transformersized to accommodate the out-of-phase ac voltage and current.
• The subsystem characteristics are summarized in the table below.
• The GEMINI CORPORATION, of MUKWONAGO, WISCONSIN, manufactures the subsystem components and WINDWORKS, INC. markets them.
KEY INVERTER DESIGN CHARACTERSTICS
OUTPUT POWER RATING
OUTPUT VOLTAGE:
INPUT VOLTAGE:
FULL LOAD POWER FACTOR:
FULL LOAD EFFICIENCY:
FULL LOAD HARMONIC DISTORTION
OPERATING TEMPERATURE
10 kVA CONTINUOUS
240 VAC Utility Residential Service
138 .:!:. 18 Vdc
60% Minimum,
92% Minimum
30% Maximum
00 to 40°C
POWER CONVERSION SUBSYSTEM
I TRANSFORMER I
- '-- + C/>A
-II ~- N
I C/>BINVERTER -
II FILTER I
LJ AND t Ii:TTERY CHARGER~
INTERLOCK
I VENTI LATION ISYSTEM 1-----.........-----'
3-33
BATTERY STORAGE SUBSYSTEM
• The battery storage subsystem provides the energy storage capacity forthe solar generated power.
• The subsystem consists of two main components: the BATTERY and theCHARGE CONTROLLER.
• The subsystem includes a 20 kWh LEAD-ACID battery made up of 64 seriesconnected cells producing a nominal 135 Vdc output.
• The charge controller maintains the voltage wi thin a 144 to 156 Vdcrange which is indicative of full battery charge.
• ESB INCORPORATED, EXIDE POWER SYSTEMS DIVISION, of YARDLEY, PENNSYLVANIA, manufactures the battery. GENERAL ELECTRIC COMPANY manufacturesthe charge controller.
102 CM (40 IN)r; II
1 2 3 4
~/1 2 rT 3 4
I~ I
r
FRONT
3-34
lh rD
c::000
fh rll
SIDE
z
::JEuU'lN-
JUNCTIONBOX ONCABINET
WIRING ISOMETRIC
91 CM (36 IN) ,-., ... ....
2 3 4
f-C n M I;;b~I
DOOR
PLAN
ARRAY/BATTERY SIZING
• The array/battery sizing uses MARGINAL COSTS and BENEFITS.
• Battery capital cost affects the relative size of the array and battery.- Battery capital costs greater than $lOO/kWh imply smaller batteries
(-20 kWh).
- Lower battery capital costs imply larger capacities for fixed arraysizes.
• System economics are sensitive to load level for all battery costs. Newoptimum system sizes must be developed for significantly varying loadlevels.
PHOENIX
O'- L..- L..-_L..- L...---'o 20 40 60 80
BATTERY CAPACITY, kWh20 40 60 80CELL AREA, m2
ATTERY COST$/kWh
15010050~o ~~
BATTERY CAP. = 20 kWh
~ 2.0U
w U..J :::. 1. 5U>-0Uj:1.0w<u.a::3 ... 0.5
(/)
oU 0 '-_"'--_"&'-_...1--'---1.._-'
o
BATTERY COST$/kWh
~15010050
- 0III
2CELL AREA = 65.0 m
-a:: 2.0U
w U..J..J1.5U->-0U-... 1.0w<u.a::3 ... 0.5
(/)
oU
ALBUQUERQUE
CELL AREA 6Sm2=
- 2.0a:: 2.0 -U BATTERY COST a::
Uw U $/kWh U 1.5..J:::' 1.5 w..JU 150
..J_>-0 UU- 1.0 ----::;::; 1 00 >-0 1.0...
:::: : ::: 5~U-
w< ...w<u.a::
0.5I u.a:: 0.5..J ... , -
(/) , ..J ...0 I (/)
U 0U
0 20 40 60 80
BATTERY CAPACITY, kWh
BATTERY CAPACITY = 20kWh
BATTERY COST$/kWh
o 20 40 60 80
CELL AREA, m2
3-35
DESIGN PERFORMANCE
• Total net system output for both Phoenix and Albuquerque is approximately 69% of the load requirements.
• Phoenix shows better load matching characteristics.
• Overall system efficiency based on incident insolation for gross arrayarea is within the 5.9 to 6.1% range for the Southwest.
UTILITY MAKE-UP, MWh - 5.0HEATING LOAD, MWh - 1.0COOLING LOAD, MWh - 11,7
2.0
NOTES
1. ARRAY CONFIGURATION1OSxl OP -65. Ii m2 CELL AREA
2. BATTERY CAPACITY 20 kWh
1.8
0.4
0.2
>U
"UJZUJ
~:I:IZo::0.6
UTILITY MAKE-UP, MWhHEATING LOAD, MWhCOOLING LOAO, MWh
4.6- 5.3- 2.8
NOTES
1. ARRAY CONFIGURATION10Sxl0P-65 m2 CELL AREA
2. BATTERY CAPACITY25 kWh
LOAD(14.8 MWh)
NETSYSTEM OUTPUT
(10.6 MWhj
EXCESS(1. 0 MWhj
2.0
1.8
1.6
~ 1.4::::
>- 1.2U
"UJZ 1.0UJ
>-...J
0.8:r:I-Z0:: 0.6
0.4
0.2
3-36
PASSIVE HOUSE DESIGN FOR THE NORTHEAST
A 4.1 kW grid-connected PV system with utility feedback is designed for a
passively heated house having low total electrical load requirements. The major
system elements are the photovol taic array and the power conversion subsystem.
Marginal life cycle cost analysis was used for system sizing. The analysis
considers various array sizes and provides parametric performance results. In
general, with utility sellback rates of 50% or greater, the results indicate a PV
array size as large as the available roof area. This design considers a low
energy consuming house with a minimum of available roof area. The 50 m2 solar
array uses a rectangular shingle solar cell module being developed by General
Electric. The array is wired in a redundant series/parallel matrix.
3-37
HOUSE DESCRIPTION
• The house design is a TWO-STORY residence with a basement of NEWCONSTRUCTION for the Northeast region of the country.
• The design includes PASSIVE SOLAR FEATURES and ENERGY CONSERVATIONFEATURES projected in 1986.
• There is 157 m2 (1690 ft 2 ) of living area with a 9 m2 (96 ft 2 )greenhouse and 57 m2 (614 ft 2) of south facing roof area.
• The house is ALL ELECTRIC with a 2 1/2-ton heat pump and electric hotwater heater.
• The site layout has a detached garage with a lot area of 1/4 acre.
3-38
SYSTEM DESCRIPTION
• The system is grid connected with a 4.1 kW NOCT array rating using a GErectangular SHINGLE MODULE ARRAY. The array consists of a total of 52FULL AND 8 HALF MODULES COVERING 50 m2 in a redundant parallel-seriesnetwork.
• The power conversion subsystem uses a 4 kVA LINE COMMUTATED MAX POWERTRACKING INVERTER to convert dc generated power to ac. A 5 kVA SINGLEPHASE ISOLATION TRANSFORMER is used to match ac supply voltage to theload.
• The system operation is PARALLEL AND SYNCHRONIZED WITH THE UTILITY.
• Excess generated power is FED BACK to the utility.
• The system represents the SIMPLEST PHOTOVOLTAIC DESIGN with a minimumof components and controls.
UTILITY FEEDBACK SYSTEMUTILITYBACK UP
3-39
SYSTEM OPERATION
• The system has AUTOMATIC STARTUP and SHUTDOWN control.
• The system automatically shuts down with loss of the utility power.
• System operation is summarized by the sequence below:
1. At sunrise, in the automatic "on" mode, the ac and dc contactorswill close when the array bus voltage reaches a threshold at lS6 Vdc.
2. During the daylight period, the inverter operates continuously ifthere is a net power output.
3. The inverter will track the maximum power operating point within +1percent over the range of 156 to 180 Vdc.
4. The interruption of utility-supplied power opens the dc contactorand it remains open until the line voltage is restored.
5. At sunset, the inverter ac and dc contactors open when the net poweroutput falls to zero. These contactors· remain open throughout thenight to eliminate the majority of the inverter parasitic losses.
unU1't_~"""S1NGLI-'-YI>C
WALL
~-
;-40
POWE" CONVlfI..ON SYSTEM sr~YIC£ PA'I!l
PHOTOVOLTAIC ARRAY
• The array consists of SHINGLE PV MODULES connected in an 8 series by 7parallel network covering 50 m2 of roof area.
• The array orientation is due south with a roof pitch of 33.70 • Theoverall cell packing efficiency is 97% over the 50 m2 •
• The modules are DIRECT MOUNTED on top of the roofing felt and plywoodroof sheathing. They form a weather tight roof.
• The modules are installed by an overlapping procedure similar to conventional shingles. Each shingle module requires two FCC butt terminationsto electrically interconnect the module to the array. Standard roofingnails are used for attachment to the roof.
3-41
,,
PHOTOVOLTAIC MODULES
• The module is currently under development by General Electric Company.
• The module uses 94 mm SQUARE SILICON CELLS. with 48 cells connected inseries and 2 parallel circuits per module.
• A half-sized module with one series string of 36 cells is also used inthe installation.
• For a NOCT of 650c, the MAXIMUM POWER OUTPUT is 74 Wand 16.6 V at SOCconditions (1 kW/m2 , 200C ambient, 1 m/s wind speed).
A summary of the module characteristics are:
Full Module Half Module
Module weight: 18 kg 9 kg
Total Cell area: 0.866 m2 0.433 m2
Exposed Module area: 0.916 m2 0.458 m2
Module packing factor 0.945 0.945
ILo.7716..--J1Lo.77481J1~
.049m- I--t- I-- 0.15..
HALF
SHINGLE
H (+)
.-r --~ ~
0.25.;
-i
I
I +I
e
I : + . i J1.37m
o 58m ±.. ~'- -L,_,-,- - ~'L II-I-l--I-+L--I±~~1- ~
I': -f-oo"o.=0==J. • 1.54..··
1.55.. )
0.1111 ..
V""~0
1.73..
O.86m
LFULL SHINGLE
,-H (+)
3-42
POWER CONVERSION SUBSYSTEM
• The PCS provides the interface between the PV array and the normalresidential utility service and loads.
• The subsystem consists of three main components: the INVERTER, the DCFILTER and TRANSFORMER along with the associated control circuitry.
• The subsystem is rated at 4.0 kW of power output with a 5 kVA transformer, sized to accommodate the out-of-phase ac voltage and current.
• The subsystem characteristics are summarized in the table below.
• The Gemini Corporation of Mukwonago, Wisconsin, manufactures the subsystem components and Windworks, Inc., markets them.
KEY INVERTER DESIGN CHARACTERSTICS
Output Power Rating
Output Voltage:
Input Voltage:
Full Load Power Factor:
Full Load Efficiency:
Full Load Harmonic Distortion
Operating Temperature
4.0 kW Continuous
240 Vac Utility Residential Service
168 + 12 Vdc
60% Minimum,
92% Minimum
30% Maximum
00 to 400 C
POWER CONVERSION SVSTEM
Vcc---+l
IOC--"'L~~:...J
3-43
ARRAY SIZING
• The array sizing uses MARGINAL COSTS and BENEFITS.
• Energy sellback price the utility is willing to pay the homeowneraffects the system sizing.
• The assumed energy escalation rate is 4% above inflation.
1.3 1.3 MADISON
1.2 1.2 SELL BACKRATIO
0.8
1.1oSell: 1.0ten8 0.9""-'u>u 08"" ."'---'
0.7
SELL BACKRATIO
~O'30.5
0.7
NOTES1. PRICE OF ELECTRICITY
$0.0832/ KWH
1.1o
§ 1.0tenou"" 0.9-'u>u
"""'-.-'
0.7
~O.30.5. 0.7
. NOTES- 1. PRICE OF ELECTRICITY
$0.0532/KWH
0.6 0.6
8020 40 60
COLLECTOR AREA. H2
OL--I---+--+---t--+---+---t---io8020 40 60
COLLECTOR AREA. ,;.
oo
3-44
DESIGN PERFORMANCE
• Total net system output for both Boston and Madison is approximately 40%of the total load requirements.
• Overall system efficiency based on incident isolation for gross arrayarea is approximately 8.3% for the Northeast.
2.2 2.2BOSTON MADISON
2.0 2.0
1.8 1.8
1.6LOAD
1.6 . 15.6 MWHLOAD
::c: 1.4 14.3 MWH 1.4~. 1.2 1.2~a:l.I.I
1.0z 1.0w
.8 .8
.6 PV SYSTEM .6 PV SYSTEMOUTPUT OUTPUT
.4 5.6 MWH .4 6.3 MWH
.2 SELLBACK .2 SELLBACK2.0 MWH 2.4 MWH
0 AtstoiNtol 0
UTILITY MAKE-UP 10.6 MWH UTILITY MAKE-UP 11.6 MWH%OF LOAD SUPPLIED %OF LOAD SUPPLIED
DIRECTLY 25.3 2 DIRECTLY 25.3INSOLATION ON ARRAY 1338 KWH/m INSOLATION ON ARRAY 1456 KWH1m2
3-45
INTEGRAL MOUNTED PV ARRAY FOR THE SOUTHEAST
A flat plate photovoltaic system with utility feedback is designed for a single
story, all-electric house in the Southeast. The two major system elements are the
photovoltaic array and the power conversion subsystem. Marginal life cycle cost
analysis was used for system sizing. The analysis considers several values of
array area and provides parametric results. The PV system is sized at 5.6 kW at
NOCT conditions. The array covers the complete roof area and is integrally
mounted in the roof. The 86 m2 solar array uses a modified Solarex inter
mediate load module in a 7-parallel by 14-series module array.
3-46
HOUSF. DESCRIPTION
• The house design is for a SINGLE-STORY residence of NEW CONSTRUCTION forthe SOUTHEAST region of the country.
• The design includes PASSIVE SOLAR and ENERGY CONSERVATION FEATURESprojected for 1986.
• There is 161 m2 (1736 ft 2 ) living area with 92.2 m2 (992 ft2 ) ofsouth facing roof area.
• The house is ALL ELECTRIC with a 3-ton heat pump and electric hot waterheater.
3-47
SYSTEM DESCRIPTION
• The system is grid connected with a 5.6 kW NOCT rating using SOLAREXBLOCK IV INTERMEDIATE LOAD MODULES incorporated into an integral mountdesign. The array consists of 98 modules with 74.5 m2 of aperture in a7-parallel by l4-series network.
• The power conversion subsystem (PCS) employs a 6 kVA-SELF COMMUTATEDMAXIMUM POWER TRACKING INVERTER to convert PV generated power to ac andto match ac supply voltage to the load.
• The system operation is PARALLEL and SYNCHRONIZED WITH THE UTILITY.
• Excess generated power is FED BACK to the utility.
UTILITYSERVICE
INVERTER
INTERIORSWITCH
3-48
SWITCI1
VARISTOR
PV CB
TYPICALBRANCH
SERVICEPANEl.
RESIDENTIALLOAD
----7
PHOTOVOLTAIC ARRAY
• The array consists of INTEGRAL MOUNTED PV modules connected. in a 14series by 7-parallel network covering 80 m2 of roof area.
• The array orientation is due south with a roof pitch of 22.60.
• The modules and mounting extrusions are mounted on the roof rafters andpurlins. No plywood sheating is used. The installation forms a weathertight roof.
• Mounting extensions are attached to the rafters with screws. The modulesare then bolted to the mounting extrusions. This compresses anelastomeric material to form a seal.
• Electrical connections are made with patented connectors and factorysupplied cables.
6d\V I.lo. ~ .1.. 2.. ~ .:i. .r... 1- ~ ~ .:.e. 11. 1k. ~ .M. !2.. .lli. J1. ~ .12.. 20
GBBBBGBBBGG8BBB888B8GDDDDDDDDDODDDDDDDDDBDDDDDDDDDDDDDDDDDDDGDDDDDDDDDDDDDDDDDDOt~~~
o"~~y~BEJ[JEJEJBBEJBBEJ EJEJEJEJEIEIEJG~MYf.--r-11,-'1'" <i: ,Q'~ ~~~E:I';>
3-49
PHOTOVOLTAIC MODULES
• The module is manufactured by SOLAREX CORP., of ROCKVILLE, MARYLAND andwas developed as part of the JPL Block IV procurement.
• The module uses Solarex 95 mm SQUARE CELLS with 36 cells connected inseries and 2 parallel circuits for a total of 72 cells per module.
• For an estimated NOCT of 71oC, the MXIMUM POWER OUTPUT is 56.9 W, and13.0 V at SOC conditions (1 kW/m2, 200 C ambient, 1 m/s wind speed).
• The design approach uses only the Solarex laminate assembly of themodule. A specially designed frame is used to mate with the integralroof extrusions.
A summary of module characteristics include:
Module weight: 15.9 kg
Total cell area: 0.650 m2
Exposed installed module area:0.815 m2
\CLOSURE STRAP
\ 126.0cm.
2x4 or 2x6PURLIN.
3-50
Module packing factor: 0.80
Nominal size: 1.26 m long by0.655 m wide
65.08cm.
ALUMINUM EXTRUSIONSUPPORTS
2x6 or 2x8 RAFTER
rOWER CONVERSION SUBSYSTEM
• The Power Conversion Subsystem (PCS) provides the interface between thePV array and the normal residential utility service and loads.
• The subsystem consists of three main components: the inverter, the dcfilter and transformer, along with the associated control circuitry,packaged in a single unit.
• The subsystem is manufactured by Abacus Controls, Inc., Sommerville, NewJersey.
• The subsystem is sized for an 6 kW POWER OUTPUT with the transformersized to accommodate the adjustment of the ac voltage and current.
• The subsystem characteristics are summarized below.
KEY INVERTER DESIGN CHARACTERSTICS
OUTPUT POWER RATING
OUTPUT VOLTAGE:
INPUT VOLTAGE:
FULL LOAD POWER FACTOR:
FULL LOAD EFFICIENCY:
FULL LOAD HARMONIC DISTORTION
6 kW CONTINUOUS
240 VAC Utility Residential Service
200 .:!:.. 40 Vdc
Unity
90% Minimum (over 25% to 88% ofoperating range)
5% Maximum
MAXIMUM (+) ~ (-) INPOWER 1---....-4~ ~--t PHASETRACKER SENSE
~AMPLITUDECONTROL
AC LINEROOF DC CONTACTORARRAY FILTER I SCR I XFMR ~nn :r- AC OUTPUTINPUT I I \1'1_
9 FAN
I - SENSE ( THERMALPI.L. SWITCH UTILITY
V - REFERENCE VOLTAGE
.3-51
ARRAY SIZING
• Energy sellback price the utility is willing to pay the homeowneraffects overall system sizing.
• Sellback rates greater than 50% imply full roof arrays.
• The array sizing uses MARGINAL COSTS and BENEFITS.
1.2
1.1
1.2
:::=1'$/"·.,1.5 1.1
~:.7I
PS/PE • .J
.5
----_.7MIAMI
SJ9.9/m2 INSTALLATION COST (1980$)
3-52
a:'-''-'.J
1.0
.9
.8
IIIIIIIIDESIGN POIHT ~.I
IIIII
.9
·.8
CHARLESTON, SC
SOLAREX IHTElU1EDIATE MODULES
20 30 40 50 60 70 80
COLLECTOR AREA _m2
DESIGN ELECTRICAL PERFORMANCE
• Net PV system output for Miami is 61% of the load requirements and thecorresponding value for Charleston is 68%.
"
• For Miami, approximately 67% of the PV system output is applied directlyto the house load while the utility receives the remaining energy. InCharleston, 53% is applied.
• The monthly net electric'al system output in Miami is always less thanthe total monthly load.
• Overall average annual PV system efficiency, based on incident insolation for the gross array area, is 7% for this design in the Southeast.The corresponding peak module efficiency assumed is 8.8% at 28 degrees C.
14$ X 7p CONFIGURATION
::c:3:1:
2.0
MIAMI
PV SYSTEMOUTPUT
2.0
1.5
1.0
0.5
CHARLESTON
PV SYSTEMOUTPUT
ANNUAL TOTALS ANNUAL TOTALS
LOADPV SYSTEM OUTPUTUTILITY l'lAKE-UPSELL BACK
14.9 MWh9.1 MWh8.8 MWh3.1 MWh
LOADPV SYSTEM OUTPUTUTILITY MAKE-UPSELL BACK
12.7 MWh8.6 MWh8.2 MWh4.1 MWh
3-53
A PV SYSTEM FOR A TEMPERATE CLIMATE
A 4.3 kW grid-connected PV system with utility feedback is designed for a house
having low space conditioning load requirements. The major system elements are
the photovoltaic array and the power conversion subsystem. Marginal life cycle
cost analysis was used for system sizing. The analysis considers various array
sizes and provides parametric performance results. In general, with utility
sellback rates of 50% or greater, the results indicate a PV array size as large
as the available roof area.
This design considers a low energy consuming house with a minimum of available
roof area. The design also provides a comparison of integral and stand-off
mounting techniques for the same array configuration. The array is mounted on the
garage roof which also is a different approach than previous designs.
The 40.3 m2 solar array uses a laminated glass/cell assembly module in an array
with four parallel circuits. Two circuits have 12 series modules and two circuits
have 13 series modules.
3-54
HOUSE DESCRIPTION
• The house design is for a SINGLE-STORY residence of NEW CONSTRUCTION fora TEMPERATE CLIMATE region.
• The house can be sited following a ZERO-LOT-LINE plan.
• The design includes PASSIVE SOLAR and ENERGY CONSERVATION featuresprojected for 1986.
• There is 142 m2 (1530 ft 2) of living space.
• The plan includes a TWO-CAR GARAGE with 44.4 m2 (477 ft2 ) of garageroof area available for mounting the solar array.
3-55
SYSTEM DESCRIPTION
The system is grid-connected with an 4.29 kW NOCT array rating using aglass laminate module assembly in an INTEGRAL and STANDOFF MOUNTINGarrangement. The array consists of 50 modules covering 40.3 m2•
The power conversion subsystem uses a 4 kVAINVERTER to convert dc generated power to acvoltage and load.
MAXIMUM POWER TRACKINGand to match ac supply
• The system operation is PARALLEL AND SYNCHRONIZED WITH THE UTILITY.
• Excess generated power is FED BACK to the utility.
• The system has automatic STARTUP AND SHUTDOWN control and shuts downwith loss of the utility.
• System operation is summarized by the sequence below:
1. At sunrise, in the automatic mode, when the input voltage exceeds180 V, the ac line contactor is closed and the phase locked loop andmaximum power tracker are energized.
2. During the daylight period, the inverter operates until the dc inputvoltage falls below 160 V.
3. The inverter tracks the maximum power point within .:1% over therange of 160 to 240 Vdc.
4. The interruption of utility supplied power opens the ac contactorand it remains open until the line voltage is restored.
5. At sunset, the inverter ac contactor opens when the net power outputfalls to zero. This contactor remains open throughout the night tominimize parasitic losses.
UTILlTVSERVICE
INVERTER
3-56
EXTERIORSWITCH
INTERIORSWITCH
SWITCh
MAINCB
VARISTOR
PVCB
TYPICALBRANCH
SERVICEPANEL
RESIDENTIALLOAD
PHOTOVOLTAIC ARRAY
• The array consists of PV modules and mounting accessories which may beused either as an INTEGRAL MOUNT or as a STANDOFF MOUNT. There are 4circuits connected in parallel with two 13 series circuits and two 12series circuits.
• The array orientation is due south with a roof pitch of 18.40 •
• The modules are mounted on a series of channel supports which can beplaced on the roof truss system for an integral mount or on the shingledroof for a standoff mount. Electrical connections are made with patentedconnectors and factory supplied cables.
• The watertight· integrity of the photovoltaic roof is assured by a simplemodule perimeter seal which uses the sloping roof surface to the maximumadvantage. An overlapping seam is used between modules to shed waterwhich runs down the roof surface.
Outlet Box
Positive circuit leads~----------
Conduit:----)~II
=
t. ___c: )(41-201 '" '20'-1011 -------1
3-57
PHOTOVOLTAIC MODULES
• The module assembly is a laminated glass superstrate/cell unit similarto the Solarex Corporation, Block IV residential module, but withsealant strips attached for mounting.
• The module uses nominal 100 mm square cells with 36 series units of 2parallel cells for a total of 72 cells per module.
• For an estimated NOCT of 710 C the maximum power output is 85.8 W at 13V at SOC conditions (1 kW/m2 , 200C ambient, 1 m/s wind speed).
• A summary of module characteristics include:
Module weight = 11.69 kg
Total cell area = 0.72 m2
Exposed module area = 0.8045 m2
Module packing factor = 0.895
3-58
ClOSURE STRIP
RAFTERS OR TRUSSMEMBERS
EDGE SEALI NG STRIP
HORIZONTAL BLOCKING
POWER CONVERSION SUBSYSTEM
• The Power Conversion Subsystem (PCS) provides the interface between thePV array and the normal residential utility service and loads.
• The subsystem consists of three main components: the inverter, the dcfilter and transformer, along with the associated control circuitry,packaged in a single unit.
• The subsystem is manufactured by Abacus Controls, Inc., Sommerville, NewJersey.
• The subsystem is sized for a 4 kW POWER OUTPUT with the transformersized to accommodate the adjustment of the ac voltage and current.
• The subsystem characteristics are summarized below.
KEY INVERTER DESIGN CHARACTERSTICS
output Power Rating
Output Voltage:
Input Volt"age:
Full Load Power Factor:
Full Load Efficiency:
Full Load Harmonic Distortion
4 kW Continuous
240 Vac Utility Residential Service
200 + 40 Vdc
Unity
90% Minimum (over 25% to 88% ofoperating range)
5% Maximum
MAXIMUM (+)rt""""1(-) INPOWER 1---""" <::... .....--t PHASE
TRACKER SENSE
AMPLITUDECONTROL
AC LINEROOF DC CONTACTORARRAY FILTER SCR XFMR I'Y"'/("\ nn :l- AC OUTPUTINPUT INVERTER uu V..lI.
I - SENSE
9 'AN( THERMAL
PLL SWITCHUTILITY
V - REFERENCE VOLTAGE
3-59
ARRAY SIZING
• The array sizing uses MARGINAL COSTS and BENEFITS.
• The system shows economic viability (LCCR<l) with sellback rates of 30%or greater. These sellback rates imply full roof arrays.
1.2
SANTA MARIA 0.7
ARRAY CONFIGURATION135 x 2P + 125 x 2P
100
GARAGEROOF AREA
40 60 80
COLLECTOR AREA - m2
~PS/PE = 0.3---,--------
20
0.2
oo
0.8
1.0
a:u 0.6~
3-60
DESIGN ELECTRICAL PERFOR'tI.ANCE
• Net PV system output for Santa Maria, CA is 67% of the load.
• Performance results indicate that 56% of the PV syste~ output is applieddirectly to the house load while the utility receives the remainingenergy.
• Overall PV system efficiency based on incident insolation for grossarray area is 9.6% for Santa Maria.
1.4 SANTA MARIA
1.2.t::: LOAD:E
1.0..>tj.c::
0.8wzw>-
0.6..J:cI-z0 0.4:E
ARRAY CONFIGURATION
0.2 13S x 2P + 12S x 2P
0J M J J A S 0 N D
ANNUAL TOTALS
LOAD 11. 3 MWhPV SYSTEM OUTPUT 7.6 MWhUTILITY MAKE-UP 7.1 MWhSELL BACK 3.4 MWh
3-61·
SECTION 4
DESIGN CONCERNS
During the design effort, several concerns were addressed. Some of the design
considerations were only pertinent to a single design, but others were applicable
to all of the designs. All of these are summarized in the following sections.
Array Sizing
The primary area available for residential PV arrays is roof area. In grid
connected applications, the system size should be selected to deliver energy to
the owner's load at the minimum life-cycle cost, including the cost of backup
utility energy. Assuming 1986 array costs of 70~/kWp and 1986 system costs, study
results indicate array areas using all of the available roof area when sellback
rates are 50% or greater. Incorporating cost effective passive house design
features and energy conservation features reduces the overall household energy
requirements and thus reduces the array requirements. This allows more flexi
bility in the architectural house design. Additional background information into
system sizing considerations is provided in Reference 1.
Roof Constraints
The physical size or available dimensions of the roof can impose several con
straints on the array configuration. This is more of a problem for direct
mounting systems where arbitrary row wiring may not be possible; however, cross
row wiring in integral or standoff mountings make the installation details more
difficult and more costly.
Rectangular areas are desirable both from an aesthetic viewpoint and from an
electrical interconnect standpoint. If modules are only connected in series up
the slant height of the roof, the system operating voltage is limited to, at a
minimum, multiples of the slant height. This implies that, for different roof
sizes using the same modules, each system results in a different operating
voltage. A transformer may be required to ultimately match house loads. The
addition of a transformer to match the loads appears more practical and less
costly than changing the specifications for the inverter package for each voltage
input. Thus, a common operating voltage range is difficult to standardize.
Similar concerns result if the modules are connected in series along the roof
length. Reference 9 addresses the problem of roof constraints in more general
terms.
4-1
Array Mounting Approach
Each of the key array mounting approaches, except rack mounting, was addressed.
It is not apparent from the designs which approach is the most cost effective
technique. Indeed, each approach has its advantages and disadvantages. It is
important to note, however, that the array mounting technique has subtle effects
on the system overall costs. Module interconnection layouts, array operating
voltage, replacement or maintenance access, and weather seals all impact the
system. Some of these cost impacts will be evaluated as Sandia obtains detailed
cost estimates for each of the developed designs. In addition, the data being
generated at the Northeast and Southwest Residential Experiment Stations (RES)
will also provide guidance in array mounting approach performance and costs.
Power Conversion SUbsystem
Several design concerns exist in the area of power conversion. Experience at the
RES's have indicated problems in startup and maximUID power tracking. The designer
must be aware of the high open circuit voltage that can exist during startup
conditions when there is a low total power output. The specifications of the unit
must accommodate these initial conditions.
It is easy to specify Power Conversion Subsystem operation with ranges wi thin a
few percent of the maximum power point; however, test data on maximum power
tracking options, has indicated difficulty in achieving these specifications.
Alternate schemes are under development, including the use of a pilot cell. This
approach is an open loop control algorithm and, although relatively simple, may
not represent the true array operating point over the lifetime of the system as
module performance is degraded due to dirt, electrical problems or shadowing.
Thus, additional development is required on these units.
Throughout the study, two primary types of power conversion units were available,
the Gemini and the Abacus Sunverter. During this time period, Abacus was also
developing a lower cost version of the currently available unit. This unit was
specified in several of the designs. After the designs were completed, however,
Abacus changed their development directions and this design version was dropped
from further development. In addition, residential sized development of power
conversion units was underway by several other firms. Optimal inverter conceptual
designs have been developed by United Technologies Corporation, Westinghouse,
General Electric and Airsearch Manufacturing Company (Division of Garrett), under
4-2
DOE contract. DECC, Delta Electronic Control Division of Helionetics, Inc. and
American Power Conversion Corporation also have hardware either presently
available or to be available in the near future. Sandia is also developing a
baseline inverter specification. Thus the state-of-the-art in power conversion
equipment will be changing over the next several years and manufacturers should
be consulted for the latest available information.
Isolation Transformer and Grounding
In many of the designs, the negative bus bar was located along the eave of the
roof. The negative array busbar located near the gutters at the eave of the roof
is subjected to a potential ice buildup from the gutter under a combination of
environmental circumstances in northern installations. Water from this ice backup
could conceivably penetrate the insulation of the busbar or a slender tool being
used to clean debris or the ice blockages could come in contact with the busbar
providing a potential to ground. A wired connection to earth ground for the
negative busbar should be provided as an additional safety precaution. This
connection creates the requirement for an isolation transformer in the power
cpnversion system. While an isolation transformer is not a mandatory system
element except for array bus grounding, its presence can provide a convenient
location for adjustments of system output to the required house service rating.
This feature provides design flexibility in array circuit layout for different
roof designs. The transformer, however, adds economic penalties both in ini tial
cost and in net power loss over the life of the system operation. Elimination of
array bus grounding and the associated isolation transformer, providing the array
output matches the house service rating, can increase the system output at
negligible decrease of safety.
Exterior Disconnect Switches
Almost all of the designs had a fused exterior switch and an interior switch. The
external switch allows to open circuit the array in the event of fires, etc. It
is a redundant switch and, in some measure, tends toward a conservative design
approach. The additional devices and installation, however, are only minor addi
tional initial economic penalties. Non-functional redundancy in this chain may be
eliminated as experience with system installations is gained and NEe codes for
photovoltaic installations are developed.
4-3
Battery Location Concerns
Location of batteries within a residence presents several concerns. Trade-offs
between an interior and exterior location for the batteries resulted in the
interior location based primarily on the requirements for maintaining environment
control and access for maintenance. The batteries were placed in the equipment
room with all the PV and conventional equipment.
The ro.om must be designed to have sufficient ventilation to remove any hydrogen
buildup, to accommodate active battery ingredients spillage and to minimize
access to the equipment from unsuspecting occupants, especially children. To
prevent a hydrogen buildup within the room, a guaranteed amount of ventilation at
all times is required to dissipate the small amounts of hydrogen and chemicals
emitted under normal operating conditions and a positive 30-60 cfm ventilation is
suggested to remove hydrogen during equalization of battery charge. These venti
lation requirements result in a slight increase in infiltration loss in the
house. To accommodate any possible spillage, a 5 em drop in the slab should be
designed around the battery racks. To minimize general access to the batteries,
they can be located in a metal enclosure. The cabinet should be heavily louvered
and have a locked door. All of these design considerations increase system costs.
As storage systems become more widespread, standard designs to resolve these
concerns may be developed.
Module Interconnection
Several of the module interconnection schemes do not currently satisfy the
National Electric Code. For example, the AMP flat conductor cable used for the
direct mount shingle designs is currently in the codes for interior, commercial
under-carpet installations only. The wire runs between battens of the direct
mount batten module also would not be classified as raintight raceways, since the
conductors are not enclosed on all sides. In addition, splice connections would
not be allowed within the batten seams. These apparent code restrictions have to
be resolved for widespread implementation.
Fire Safety
Fire safety is an important issue that could not be specifically addressed in the
design other than for general considerations. It is likely that the module
installations will be considered as a roof covering by local building code
4-4
a fire rating
"Test for Fire
officials. In this case, the module installation must receive
classification in conformance with the requirements of UL-790,
Resistance of Roof Covering Materials".
Local fire companies would have to be alerted and instructed in regards to the
potential danger of the energy-producing roof. The location of the de power line
adjacent to the normal utility line and the exterior disconnect switch should
alert fire personnel that the house has an additional power supply. However, fire
personnel must be informed that disconnect at the switch still leaves the roof as
an active generator in the daylight and only interrupts power supply to the house.
4-5
SECTION 5
DESIGN STUDIES
This section discusses some design studies which were applicable to the designs
in general or were completed after the respective report was written. For
example, power conversion subsystem requirements were evaluated relative to
operating voltage ranges and losses associated with restricted voltage operation
ranges through system simulations. A correlation for estimating power losses for
limited system operation within specific voltage ranges was developed. In
addition, annual distributions of array current characteristics were developed.
Power Conversion Subsystem Correlations
During system analyses of the detailed designs, narrowing the maximum power
tracking operational voltage range over which the power conditioner operates was
found to have only small effects on the total annual energy collected. In fact,
single voltage operation in some cases might be implied. However, the amount of
lost energy and voltage point of minimum energy loss vary with several
parameters. Thus, attempts to correlate the data were made.
Initially, a series of curves showing the difference in energy loss (EV) (for a
restricted voltage range) from the maximum available (maximum power tracking
wi thout limitations, EMP) versus operating voltages were generated. Figure 5-1
shows a typical curve for Miami. The percent energy loss was determined from the
expression EMP - EV which is the ordinate in the figure.EMF
One curve represents losses incurred when the array voltage is limited on the
lower end of the voltage range from 180V to 220V. This means that the system was
constrained to operate at maximum power up to a specific voltage and then limited
to that voltage if the maximum power point was above it. Therefore, as the range
of maximum power voltage operation is increased, the loss of collected energy
over that collected for unconstrained maximum power tracking decreases. The
second curve demonstrates the effect of limiting the voltage on the high end from
220V to l80V. A minimum of three simulation runs were required to establish each
curve, and the point of intersection denotes the minimum loss voltage. Plots of
this type were generated for several locations, and provided an adequate data
base to attempt correlations. It was intended that such correlations would
5-1
LIMIT ON HIGH ENDOF VOLTAGE RANGE
LIMIT ON LOW ENDOF VOLT AGE RANGE
t::N = VOLTAGE VARIATIONABOUT MINIMUM LOSSVOLTAGE
MINIMUMLOSS
VOLTAGE195.5V
NOTES:
1. SOLAREX INTERMEDIATEPV MODULES 14S X 7P
(jZ 1··l.:U<e:l-e:w~
~:;;::::l~X<:;;..J..J::::l 10·u.:;;oe:u.VlVlo..J -3..J 10 <::::lZZ<
180 190 200 210 220
VOLTAGE LIMIT - VOLTS
Figure 5-1. System Losses for the Southeast Designas a Function of Limiting Voltage Range
establish the minimum loss voltage and corresponding losses and thereby eliminate
the need to generate six . computer runs. Note that a voltage range, V, can be
defined around the minimum loss point.
Figures 5-2 through 5-6 show similar curves for different locations and different
array configurations. Several types of panels were simulated utilizing vendor IV
characteristics, including the GE Block IV and rectangular shingle designs, the
ARCO-SOLAR Block IV batten module and the Solarex Block IV module. The locations
for the analyses were Boston, Phoenix, Santa Maria, Miami and El Paso.
Available simulation data of the residential PV systems was screened to establish
correlating parameters. The NOCT temperature and voltage at NOCT conditions were
obtained by modelling, or from vendor data. With this information, attempts were
first made to correlate minimum loss voltage, normalized by the NOCT voltage with
insolation. Figure 5-7 shows relatively good correlation between the minimum loss
voltage (VML ) normalized to the NOCT voltage (VNOCT ) with the total insola
tion (I) incident on the module, normalized to a reference insolation (IREF ).
The worst deviation is 5% from the average curve fit.
5-2
LIMIT ON LOW ENDOF VOLTAGE RANGE
NOTES:
1. ARRAY CONFIGURATION 25S X 19P,BLK IV SHINGLE
2. ALL-ELECTRIC FEEDBACK
LIMIT ON HIGHEND OF
VOLTAGERANGE
IIIIIII
MINIMUM LOSSVOLTAGE
214V
IO. OJ L-_--L__...L.__..l-..I-.......JL..-_-l.__...1.._---I
oz~
u<a::I-a:: 10.0w;;:
a:X
~oJoJ:::lu..::Eoa::u..VIVI
9 0.1oJ<:::lZZ<
190 200 210 220 230 240
VOLTAGE LIMIT, VOLTS DC
250
Figure 5-2. System Losses for Boston, Design 1.lOt
X<::EoJoJ:::lu..::Eoe: 1. 0VIVI
9oJ<:::lZZ<
LIMIT ON LOW ENDOF VOLTAGE RANGE
NOTES: _
1. ARRAY CONFIGURATION255 X 19P, BLK IV SH INGLE
2. ALL ELECTRIC, FEEDBACK
0.11.----L--...L.--..l-_..lJL..-_-l._---J180 190 200 210 220 230
VOLTAGE LIMIT, VOLTS DC
Figure 5-3. System Losses for Phoenix, Design 1.
5-3
100
~ 10::.:U<e::l-e::w3~ 1.0
X<::;;-l-l::>u.::;; 0.1oe::u.IIIIII
9-l
~ 0.01zz<
=
MINIMUM LOSSVOLTAGE
156V
LIMIT ON LOW END OFVOLTAGE RANGE
LIMIT ON HIGH END OFVOLTAGE RANGE
NOTES:
1. ARRAY CONFIGURATION ONRECTANGULAR SHINGLEMODULE 8S X 7P
5-4
o. 001 r__-l-__..J-__L....._~__..J
150 .160 170 180 190
VOLTAGE LIMIT, VOLTS DC
Figure 5-4. System Losses for Boston, Design 4.10
2
\LIMIT ON LOW ENDOF VOLTAGE RANGE
LIMIT ON HIGH ENDOF VOLTAGE RANGE
NOTES
1. 12S x 4P LAMINATED MODULEARRAY
MINIMUM LOSSVOLTAGE = 172V
10- 1 '--_~__.......-'-_.......__l--_-'
160 170 180 190 200
VOLTAGE LIMIT, VOLTS DC
Figure 5-5. System Losses for Santa Maria, Design 6.
100
""(:)z:.::u«
,f 0<:I-
0<:w:c~ 10.0
X«::;;oJoJ:Ju.::;;00<:u.Vl 1.0Vl0oJ
oJ«:JZZ«
LIMIT ONHIGH ENDOF VOLTAGE RANGE
MINIMUMLOSS
VOLTAGE197V
LIMIT ON LOW ENDOF VOLTAGE RANGE
NOTES:
1. ALL ELECTRICALFEEDBACK SYSTEMBLOCK IV SHINGLE25S X 15P
0.1 L-._.L.._-'-_.L-J'--_....._ ........_ ......._--'
180 190 200 210 220 230 240
VOLTAGE LIMIT. VOLTS DC
Figure 5-6. System Losses for El Paso Single Family
1.3
BOSTON RECTANGULAR SHINGLE 85 x 7p
125 X lip
6
----0--IREF = 6000 W-hr/m2-DAY
EL PASO BLK IV 255 x 111p
PHOENIX BLK IV 255 x 19p
MIAMI SOLAREX MODULE 145 x 7p
SANTA MARIA LAMINATED ARRAY
o BOSTON BLK IV 255 X 15p
o6
o~
•
1.2
1.1
1.0
0.9
O. 8 OL.~6---0-.1-7----0.....-8---0.....-9----1·.0----1.·1----1..J.2
I "REF
Figure 5-7. Comparison Between Minimum Loss Voltage and Insolation.
5-5
Figure 5-8 shows the correlation of the same voltage ratio (VML/VNOCT) with
average ambient temperature normalized to the NOCT temperature (TNOCT
). Again.,
a relatively good correlation is achieved with a 3% deviation of the average fit.
Once the minimum loss voltage was determined, the energy loss effects of con
straining the maximum power tracking range were correlated. The correlating
parameter was the ratio of the energy loss from full maximum power tracking for
a I:1V operating range (~p - EI:1V) to the maximum energy loss at the minimum
loss voltage (EMP - EML ). This ratio ranges from 0 to 1 and is plotted versus
the normalized contained voltage operating range (AV/VNOCT ). Figure 5-9 shows
the correlation.
An example illustrates the use of these correlations to determine the fixed
vol tage minimum energy loss point, VML , and the constrained operating voltage
energy loss, EAV • Assume the following:
Location: Phoenix, I = 6560 wh/m2-day
Array Area: 93 m2
183 V
10 V
For these assumptions, the parameter I/IREF is 1.09. Using Figure 5-7, the
ratio (VML/VNOCT) is 1.04. Thus, the operational voltage point for minimum
energy loss from full maximum power tracking for a Phoenix location is 190V.
Computer simulations can be made to establish the annual collected energy at full
maximum power tracking,
lected at the fixed
example).
EMP ' (17.5 MWh for this example) and the energy col
voltage minimum loss point, EML (17.15 MWh for this
Thus, for a constrained operating voltage range of + V
AVloss 'voltage point = 0.055 and Figure 5-9 givesVNOCT
10 V around the minimum
(EMP - EI:1V)
(EMF - EML )= 0.78
Solving for EAV provides the estimate of the energy loss from full maximum
power tracking of 17.23 MWh or a 1.5% energy loss.
5-6
1.3
1.2
1.1
1.0
0.9
o BOSTON BlK IV 25S X 15P
o BOSTON RECTANGULAR SHINGLE
6. El PASO BlK IV 25S X 14P
o PHOENIX BlK IV 25S X 19P
~ MIAMI SOlAREX MODULE 14S X 7P
8S X 7P
0.80.8 0.84 0.86
TAMSTNOCT(OK) I(OK)
0.90
Figure 5-8. Correlation Between Minimum Logs Voltage andNormalized Ambient Temperature
I I
.15.le
o BOSTON BlK IV 255 X 15P
9 BOSTON RECTANGULAR SHINGLE 85 X 7P
6. El PASO BlK IV 255 X 14P
o PHOENIX BlK IV 255 X 19P
b. MIAMI SOlAREX MODULE 145 X 7P
• SANTA MARIA lAMINATED ARRAY 125 X 4P
.05
0.2
0.6
0.4
0.8
I(V~:cT)1Figure 5-9. Correlation Between Normalized %Increase in Loss with
Normalized Voltage Operating Range about Minimum Loss Voltage.
5-7
Sensitivity of Photovoltaic Module Performance to Roof Insulation for theIntegral Mount Configuration
A parametric study was conducted to determine the performance sensitivity of an
integrally mounted photovoltaic module to roof insulation and ambient conditions.
In the analysis, an attic temperature profile was assumed as two piecewise
continuous linear functions of ambient temperature. Results generated using a
one-dimensional Fourier conduction approach indicate that the backface insulation
has little ~5%) effect on module performance, and the highest module efficiency
is achieved for the case of no insulation.
The cell temperature has a significant impact on PV module performance, and this
necessitates the use of an accurate technique for predicting collector heat loss.
Reference 10 provided equations for radiative heat loss while the convective heat
loss relations were furnished by Reference 11. A simple one-dimensional Fourier
approach provided a means of establishing the effective conductance from the
solar cell to the attic. A cross section of the PV module is illustrated in
Figure 5-10 which shows this thermal path, and Table 5-1 presents the correspond
ing material and geometric properties. An energy balance was written for the
module in terms of the front face and back face heat losses, and a Newton-Raphson
method proVided a solution for cell temperature.
The backface heat loss is expressed as:
Qb = Ceff (TCELL - Tattic) (1)
where Ceff is the effective conductance from the cell to the attic and Tattic
is the attic temperature. Two piecewise continuous linear profiles, as shown in
Figure 5-11, were assumed for Tattic as a function" of outdoor ambient tem
perature.
The net front face heat transfer expression is:
+ 3·81 • VWIND
where QABS is the amount of solar energy absorbed by the collector and er isthe collector tilt angle with respect to the horizontal.
5-8
OUTDOOR AIR
ALUMINUM BACK PLATE
GLASS
EVA "-
CELLEVA,
"-
/ TEDLAR .;
1 AIR GAP
/ / /'! - I / /, /' , , /
-(
/ I rI I I- POL YURET1fANE / -
, INSULATION/ i (
(/' / ( -
-,
/( / /" ./, ---
I AIR GAP
~
.'
ATTIC AIR
Figure 5-10. Module Cross Section for Integral Mount Configuration
Table 5-1. Material Combinations for Thermal Path from Solar Cell to Attic
Thickness h k RMaterial (mm) (W/m2_oC) (W/m2_oC) (m2_oC)/W
. -_ .• ------ _.• -.•.•.c::::::::::.....
EVA 0.508 1.64 0.001
TEDLAR 0.051 5.68 0.475 0.00035
AIR GAP 2.54 0.063
POLYURETHANE 25.4 1.08 0.77INSULATION
AIR GAP 2.54 5.68 0.063
I ALUMINUM 1.016 750 5.2 x 10-6
!BACK PLATE
AIR-NATURALI 9·1 O.ll
I CONVECTION
5-9
The energy balance is:
Qb - Qf = 0 (3)
which provides a steady state solution for TCELL '
Figure 5-12 shows the effects of wind velocity and ambient temperature upon cell
temperature and module efficiency for 1 kW/m2 incident and R-4 insulation. Both
ambient temperature and wind speed are observed to have a significant effect.
The sensitivi ty of cell temperature and efficiency to roof insulation is pre
sented in Figure 5-13. The amount of insulation varies from 0 to R-26 (approxi
mately 6" thick), and the greatest influence occurs as the wind velocity de
creases to 0 m/s. The effect on efficiency is less than a 5% reduction as R is
increased, and very little effect is observed at the higher wind speeds. This
indicates that the front face heat loss is more dominant than that of the back
face for the integral mount configuration.
The results of the analysis established the following conclusions:
1. Roof insulation affected module performance by less than 5%.2. Front face heat loss was more significant than backface losses.
3. Wind and ambient temperature may affect performance by as much as 20%.
This analysis can also apply to a direct mounted array with insulation in the
attic ceiling.
Alternate Battery System Shunt Analysis
The design of the battery system in the Southwest was developed so that when the
battery was fully charged and the load was supplied, waste energy would be
shunted to ground. A possible option of utilizing this energy was to direct it to
a resistive heating element in the domestic hot water tank, thus, using the hot
water tank as an additional storage element. In effect, however, as energy is
directed and stored in the hot water tank, the total house load is reduced and.
thus, additional waste energy is generated. A quantitative evaluation of usable
energy was performed based on the system analysis results presented in Reference
5. The design analysis predicted an excess energy of 1.0 MWh in Albuquerque for
an array configuration of 65 m2 of active cell area and a battery capacity of
25 kWh. Almost all of this occurs in the spring and fall seasons. In order to
establish how much of this energy could be utilized, typical hourly hot water
demand profiles used in the analysis were examined. The hourly profiles of
5-10
TATTIC = 2.667 T AMB - 28.33°C
T ATTIC = TAMB + SoC
20
60
40
80
100
UlI-
...<
.....U
°~
00 10 20 30 40 50
TAMB - (DC)
Figure 5-11- Assumed Attic Temperature as a Function of Ambient Temperature
120 0.058
0.066
0.098
0.074>-UZW
U u0 ii:-...I 0.082 LL
...I WW WU ...I
f- :::>c0::E
40 0.090
l.-__--L ....L... .L.-__--J 0.106
10 20 30
Figure 5-12. Cell Temperature and Module Efficiency as a Function ofAmbient Conditions for the Integral Mount Configuration
5-11
120
lKW/M2
INCIDENT INSOLATIONTATTIC= TAMB + 5°C (TAMB~ 200 cl
TATTIC= 2.667 TAMB - 28.33°C (T AMB ;;'20 0 C)
0.058
100
80
;.'60
-'-'wU...
40
20
PANEL BACKFACEPOLYURETHANCE INSUL.
- R-26
___ R-4
NO INSULATION
0.066
0.074,.UZwU
0.082 u:u.ww-'::JC0:;;
0.090
0.098
L-__--'----__-I- L-__I 0.106
20 40 60 80
T ATTIC (OCl
Figure 5-13. Cell Temperature and Module Efficiency as a Function ofAttic Temperature for the Integral Mount Configuration
available excess energy were also constructed and superimpos,ed on the demand
profiles. Figure 5-14(a) illustrates the amount of PV array excess energy
available for a typical spring day. Note that the battery is not fully charged
until noontime. Between noon and 5 0' clock, there are 16.5 kWh of excess energy
present and approximately 16.0 kWh are effectively transferred to a pre-heat
water tank. A two-tank system was considered for simplicity in this analysis. The
corresponding hot water usage is shown in Figure 5-14( b). The hourly pre-heat
tank temperature is shown in Figure 5-14(c). As the excess energy is input, the
temperature steadily rises until the maximum tank temperature of 76.7°C is
reached. This stored thermal energy is sufficient to supply a substantial portion
of the hot water load during the evening peak hours and beyond. On a daily basis
in springtime, 47% of the hot water load can be provided by usage of available
excess energy.
Similar results exist for the fall season. If all of the energy transferred to
the hot water system is used prior to the start of excess energy availability,
5-12
6 ... NOTES: r--
1. ARRAY CONFIGURATION5 ~ lOS X lOP EXCESS ENERGY
~
0::: 4 65 m2 CELL AREA 16.5 kWhW I-3 2. BATTERY CAPACITY0-- 20 KWH ....-Q,J: 3 l-(/)3(/):ll::w ....
2U l-XW
1 I-
0 • • I • • 11. t t •(A)
12
W 100<
8(/)U::J--- 0:::1-0:::J:(/)w-' 6wl-.J:E«0 3 0 4o ....
I-0
2J:
·0
... ....-
l- I--
r--I- - ..... - I--
~
- ~ -~.....I- - ~
-n -t • r
(B)
,. r--I--~-
I-~
--I<- r-- ~ --...~l-
I<- ....-
~
~ • • • • •
75
~ZW 65<0:::I-::J
1---I-<U 55< 0::: 0ww ....J:Q, 45I:EWw0:::1- 35Q,
25
15M 2 4 6 8 10 N
(e)
2 4 6 8 10 M
Figure 5-14. Available Power, Hot Water Usage, and Preheat Tank Temperaturefor a PV System in Albuquerque
5-13
the next day no additional waste energy is generated. If, however, the hot water
storage displaces loads during PV system output, additional waste energy may be
generated.
Overall, this option does provide a means of utilizing waste energy.
System Current Characteristics
In all of the designs, annual distribution of the array voltage and power charac
teristics were generated to predict power conversion subsystem requirements. The
current characteristics were not, however, usually reviewed. It is feasible that
under certain weather conditions, array current output could be above power
conversion equipment limits. Thus the annual distribution of current was
calculated and plotted for several designs. Figure 5-15 shows the hours of
operation for current level for the sixth system design in Santa Maria. A peak
current output of 26 Amps is noted. Figures 5-16 and 5-17 shows similar curves
for the fourth design in Boston and Madison. Typical annual current distributions
at or below system power levels were also determined for the fourth design as
evidenced in Figures 5-18 and 5-19 for Boston and Madison. These curves defined
maximum current requirements of power conditioning equipment and estimate the
amount of energy output occuring at the current levels.
5-14
28
26 AMPS
5000
4010 HOURS
ARRAY CONFIGURATION13S x 2P + 12S x 2PINTEGRATEDARRAY MODULES
2000 3000 4000
HOURS AT CURRENT> I
1000
OL.-__...l-__--J. .L.-__-1..__---J
o
Ul0.
~ 16
....Zw0::0::B 12
20
Figure 5-15. Current Duration for the Sixth Design, Santa Maria, CA.
50
3500300025002000
NOTES:
1. ARRAY CONFIGURATION 56RECTANGULAR SHINGLES 8S X 7P
15001000500o ......--~--....J-----L.----L----.JL...---..I...,;;=-----Io
40
~ -_. 33. 6 AMPS
~30
l
f- 20zLU0:::
~ 10U
HOURS AT CURRENT> I
Figure 5-16. Current Distribution for the Fourth Design, Boston
5-15
50
40
l/)wffi 30c..:E0«I-
ffi 200::0::::::>U
10
NOTES:
1. ARRAY CONFIGURATION56 RECTANGULAR SHINGLES8S X 7P
3580 HOURS
o 500 1000 1500 2000 2500 3000 3500 4000
HOURS AT CURRENT> I
Figure 5-17. Current Duration for the Fourth Design PV Array, Madison
NOTES:
1. ARRAY CONFIGURATION 56RECTANGULAR SHINGLES8S X 7P
so40
33.6 AMP
1_- 6• 4 MWH
20 30
CURRENT, AMPERES
10
7)-
(,J::x:
ffi:: 6z~w ....lo-
S<.V::::>...lZwz> 4 -<'w)-...l<'0:::O:::w 3 -0:::::<'0
I0:::0- I
<'1- 2 f-...l<.0l/l
1 I
00
Figure 5-18. Current Distribution as a Function of Solar Array Energy for Boston
5-16
594010
0..--.---.........----.&..- --'- ...1..- -1
o
8 33.4 AMP
I 7.1 MWH7
>-(J:t
ffi:: 6NOTES:z:E
w , 1- ARRAY CONFIGURATION 56...JQ. 5 RECTANGULAR SHINGLES<V:J...J 8S X 7PZwZ> 4<w>-...J<0::C:::W 30::::<0c:::Q.
2<I-...J<0V)
Figure 5-19. Current Distribution as a Function of Solar Array Energy, Madison
Array Open Circuit Voltage
Many of the designs developed used the array open circuit voltage as the
condition for start-up. Some early test data from the Northeast and Southwest
Residential Experiment Stations indicated that actual open circuit voltages could
exceed power conversion equipment limits under certain weather conditions. Thus
for the temperate climate design in Santa Maria, Reference 8, the open circuit
voltage was plotted for several weather conditions. Figure 5-20 shows these
results. The array NOCT voltage is 185V. A curve similar to this figure helps to
identify maximum array open circuit voltage that may exist for a given array
configuration and should be investigated to assure system operation within the
power conversion subsystem limitations.
5-17
300ARRAY CONFIGURATION: 12S X 4P LAMINATED ARRAY MODULE IN SANTA MARIA
0-50100
TOTAL NORMAL
INTENSITY (W 1M 2)
50
V = 0 MIS
WIND SPEED
V = 1 MIS
403020
T AMB - °C
10o
........................
......................
..........................
......................
.................. , ......
..........................
............ 1500
240
260
200 II...- --' --L ~____.,;;::t.__L. ...DI> ___J
-10
U>0 220
wl.J«I-1o>I;:)
U~
U
ZWCO
VlI-1o 280>
Figure 5-20. Effects of Ambient Temperature and SolarIntensity on Open Circuit Voltage
5-18
SECTION 6
REFERENCES
1. "Photovoltaic System Sizing Analysis," G. Jones & E. Mehalick, Paper Presented at the Fourteenth IEEE Photovoltaic Specialists Conference, SanDiego, CA, January 7-10, 1980
2. "Final Report - Regional Conceptual Design and Analysis of ResidentialPhotovoltaic Systems," SAND78-7039 , General Ele~tric Company, January, 1979
3. The Design of a Photovoltaic System for a Southwest All-Electric Residence,"SAND70-7056, General Electric Company, February, 1980
4. The Design of a Side-by-Side, Photovoltaic Thermal System for a NortheastAll-Electric Residence," SANDSO-7148, General Electric Company, November,1980
5. The Design of a Photovoltaic System with On-Site Storage for a SouthwestAll-Electric Residence, SANDSO-7170, General Electric Company, March, 1981
6. "The Design of a Photovoltaic System for a Passive Design Northeast AllElectric Residence," SANDSO-7171, General Electric Company, October, 1981
7. "The Design on a Photovoltaic System for a Southeast All-ElectricResidence", SANDSO-7172, General Electric Company, October, 1981
8. "The Design of a Photovoltaic System for a Temperate Climate All-ElectricResidence," SANDSO-7l73, General Electric Company, December, 1981
9. "Integrated Residential Photovoltaic Array Development," DOE/JPL 955894-3,General Electric Company, August, 1981
10. "Solar Energy Thermal Processes," J.A. Duffie and W.A. Beckman, Wiley 1974
11. "Thermal Performance Testing and Analysis of Photovoltaic Modules in NaturalSunlight," LSSA Project Task Report 5101-31, Jet Propulsion Laboratory,California Institute of Technology, 29 July, 1977
12. G. Darkazalli, "A Description of the University of Texas at Arlington SolarEnergy Research Facility Photovol taic/Thermal Residential System,"MIT-Lincoln Laboratory, COO-4577-5, March 16, 1979
13. E.C. Kern and M.C. Russell, "Hybrid Photovoltaic/Thermal Solar EnergySystems," MIT-Lincoln Laboratory, COO-4577-l, March 27, 1978
14. E.C. Kern and M.C. Russell, "Optimization of Photovoltaic/Thermal CollectorHeat Pump Systems," MIT-Lincoln Laboratory, COO-4577-7, June 1979
15. "Conceptual Design and Analysis of Photovoltaic Systems," Final Report,General Electric Co., ALO-3686-14, March 19, 1979
6-1
16. E.J. Buerger, et aI, "Regional Conceptual Design and Analysis Studies forResidential Photovoltaic Systems," General Electric Co., SAND 78-7039,January, 1979
17. N.F. Shepard, R. Landes and W. Kornrumpf, "Definition Study for Photovoltaic Residential Prototype System," General Electric Co., ERDA/NASA 1979/76-1, September, 1976
18. P.F. Pittman, "Conceptual Design and Systems Analysis of Photovoltaic PowerSystems," Final Report, Westinghouse Corp., ALO/2744-13, May, 1977
19. P.F. Pittman, et aI, "Regional Conceptual Design and Analysis Studies forResidential Photovoltaic Systems," Westinghouse Corp., SAND 78-7040, July1979
20. "Photovoltaic Systems Concept Study," Spectrolab, Inc., ALO/2748-12, April,1977
aI, "Definition Study for Photovoltaic ResidentialFinal Report, Martin Marietta Corp., ERDA/NASA-19768,
M.S. Imamura, etPrototype System,"September, 1976
22. V. Chobotov and B. Siegel, "Analysis of Photovoltaic Total Energy Systemsfor Single Family Residential Applications," Aerospace Corp. , ATR-78(7694-02)-1, August, 1978
21.
23. P.R. Carpenter and G.A. Taylor, "An Economic Analysis of Grid-ConnectedResidential Solar Photovoltaic Power Systems," MIT Energy Laboratory, MTI-EL78-007, December, 1978
24. Kern, E.C., "Phase-One Experiment Test Plan Solar Photovoltaic/Thermal Residential Experiment," COO-4577-6, March 15, 1979
25. Final Report - Analysis and Design of Residential Load Centers," SAND80-7017, General Electric Company, October, 1981
6-2
APPENDIX ASUMMARY OF SYSTEM
CONFIGURATION EVALUATIONS
APPENDIX A
SUMMARY OF SYSTEM CONFIGURATION EVALUATIONS
PV ONLY SYSTEMS
System I(a) All Electric System/Battery Storage
Block Diagram
BACK-UP
PV DC!AC GENERALARRAY INVERTER LOADS
~_....SHUNT - HEAT- PUMP
TIHOT
"""'-WATER
Description
This system involves all-electric loads with space heating and cooling providedby a heat pump. The PV system includes batteries connected directly across thesolar array bus, with the solar array operating point established by the voltageof the battery. Battery charge control is accomplished by limiting the batterycharge voltage to a prescribed level. Control is exerted by partial shunting ofdiscrete sections of the solar array current to limit battery charge voltage.Excess battery temperature is also limited by shunting out array sections. Batterydischarge control is only exerted when the battery has been almost depleted, asindicated by a lower discharge voltage limit. When this occurs, the transfer ofpower through the inverter is interrupted completely and is not restored untilthe battery is fully charged. This would occur when the upper charqe voltaqelimit is reached. The inverter supplies power up to its rated limit. Any demandin excess of this limit is supplied by direct tie-in to the utility.
Variations of this basic system involves, (1) utilization of excess PV energy foruseful auxi1iarv heatina (e.g., domestic hot water) in place of wasteful dissipationthrough the shunt and (2) nightime utility charging of the batteries where advantage can be taken of lower nightime rates.
A-l
Studied By
• General Electric (References #15, #16, and #17)
• Martin Marietta (Reference #21)
, Westinghouse (Reference #18)
• MIT-Lincoln Lab. (Reference #14)
Advantages and Disadvantages
Advantages
• No max power tracker
• Does not require feedback to utility
Conclusions
Disadvantages
• Initial cost and maintenance of batteries
• Energy dissipated throughshunt when PV system output exceeds what load andbattery will accept
• Energy losses through batteryheat dissipation
• Digital shunt and batterycharge control devicesrequired
• Somewhat lower annual energydisplacement than no batterysystem
• Battery generated hydrogen gasexplosive harard. Cost ofsafety features to meet localcodes
This PV only system is less economical and displaces less energy than a systemwithout battery storage utilizing a maximum power tracker and utility feedback.Major factors contributing to disadvantage of this system include cost of batterystorage and the losses associated with the charge/discharge cycles.
Though applicable to all regions, this system proved most cost effective in highinsolation areas, providing a significantly larger percentage of electrical load,for-example, in Phoenix than in Boston. A considerable battery development effortis of course necessary to meet battery projected 1986 cost and technical goals.
A-2
System I(b) All Electric/Feedback
Block Diagram
UTILITY
PV DC!AC GENERALARRAY INVERTER LOADS
MAX. POWERTRACK
HEAT~
PUMP
- HOTWATER
Description
This system involves all-electric loads, wlth space heating and cooling provided by a heat pump. The PV system includes a maximum power tracking inverterand permits feedback of excess PV energy to the utility grid connected in parallel with the photovoltaic system. The utility distribution system essentiallybecomes the storage medium for the residential PV system. When PV maximum poweroutput is less than load demand or during nightime periods, power is supplied bythe utility.
Since the inverter AC power output is synchronized with the utility back-up source,any loss of utility power results in interruption of the inverter operation. Thisinterruption is necessary to prevent excess array output from damaging residentialloads through over-voltage.
Studied By
o General Electric (References #15, #16, and #17)
o Martin Marietta (Reference #21)
• Aerospace (Reference #22)
A-3
Conclusions
Advantages
• Simplest system
• Utility essentiallyacts as electricalstorage medium forall excess energygenerated by PV array
Conclusions
Disasvantages
• Acceptance of feedbackby util ity
• Low buy-back rate makessystem less attractive
This system is the least complex to implement, assuming of course, utilityacceptance of excess PV array output. Large scale sell-back of power tothe utility from many residences can pose a serious problem to the utility.
Of the PV only systems investigated, the feedback system provided the highestenergy displacement for an all-electric residence. Besides being approproatefor all regions, it provided higher performance and cost effectiveness than PVonly systems with batteries.
System I(c) Fossil Heating/Feedback
Block Diagram
BACK-UP
PV DC!AC GENERALARRAY INVERTER LOADS
MAX. POWERTRACKER VAPOR
""'- COMPR.COOLING
FOSSILFUEL FURN.
HEATING
HOTWATER
A-4
Description
This system involves the general household electrical loads and an electricallydriven vapor compression space cooling unit. Domestic hot water and spaceheating are provided by a fossil fuel fired furnace. This PV system is identical to system Ib, having a maximum power tracking inverter and feedback ofexcess energy to the utility grid.
Studied By
• General Electric (Reference #16)
• Aerospace (Reference #22)
Advantages and Disadvantages
Advantages
• Utility essentiallyacts as electricalstorage medium
• Simpl~ system
Conclusions
Disadvantages
• Acceptance of feedback by utility
• Economic viabilitydependent on buy-backrate
• PV cannot be useddirectly for heating
This system proved only slightly less economical than the PV only feedbacksystem lb. Thus, either back-up energy form, electrical or fossil, can effectively be accommodated by the PV only systems.
Fossil fuel fired burners providing space heating and hot water are primarilyapplicable in the cold and moderate (hot/cold) climate regions where they arepredominantly in use today.
A-5
System led) All Electric/Battery Storage/Max. Power Tracking
Block Diagram
BACK·UP
PV POWER DC/AC GENERAL...ARRAY TRACKER INVERTER LOADS
~~-- -- HEAT, PUMP
I..- HOTWATER
Description
This system involves an all-electric load, with space heating and cooling provided by a heat pump. The PV system includes a pulse width modulated (PWM) downconverter/maximum power tracker in series with the inverter, and battery storageunder control of a battery charge controller. The PV system operates at maximum power, unless the output exceeds the load and battery requirements. At thispoint, the duty cycle of the PWM automatically decreases to move the systemoperating voltage toward the open circuit voltage of the array until the available source power is equal to the total power demand of the load and batteries.The utility tie-in permits grid power to supplement PV maximum power output whenload demands exceed the photovoltaic system output.
A variation of this system is the placement of the pulse width modulatedconverter/maximum power tracker in the charge leg of the batteries (i.e., inparallel with the inverter). Therefore, PV power can pass directly to the inverter and on to the load without passing through the PWM/maxim4m tracker. OnlyPV power to the batteries passes through the PWM/maximum power tracker. Withthis system, maximum power tracking takes place only during periods of batterycharging. During the sunrise and sunset periods of the day, when some loadsharing battery discharge is required, the solar array bus voltage is clampedto the battery discharge voltage and the maximum power tracking battery chargeregulator is disabled. Maximum power tracking takes place when array output
A-6
capability exceeds the load demand. However, as with the series system, ifthe total available power at the array maximum power point is greater thanwhat the inverter and batteries will accept, the duty cycle of the PWM regulator automatically decreases to move the system operating voltage toward theopen circuit voltage of the array until the available sourqe power is equalto the total power demand of the load and batteries. When the load demandexceeds the power available from the solar array, the batteries discharge, andas previously indicated clamps the DC bus voltage to the battery dischargevoltage, forcing operation at a voltage considerably less than the solararray maximum power voltage.
Studies By
• General Electric (Reference #17)
• Martin Marietta (Reference #21)
Advantages and Disadvantages
Advantages
o Potentially highersystem output withpower tracker forbattery storagesystems
• Does not requirefeedback to utility
Conclusions
Disadvantages
; Increased controlcomplexity
~ Cost of additionalequipment
~ Higher losses
These systems with battery storage and maximum power tracking, are higher costconfigurations and displace less energy than either the battery/shunt system(la) or maximum power tracking/feedback system (lb). Therefore, they provide nosignificant improvement or advantage over these other PV only systems.
Comparison Evaluation of PV Only Systems
A comparitive evaluation of each of the PV only systems against a set of criteriapreviously described in Section 2.0,is presented in Table A-I. The evaluationresulted in the following rankings of PV only systems.
Rank
1
23
4
System Title
Ib, All Electric/FeedbackIa, All Electric/BatteryIc, Fossil Heating/FeedbackId, All Electric/Max. PowerTracking/Battery
A-7
> I m
TABL
EA
-I.
COM
PARI
SON
EVAL
UATI
ONOF
PV-O
NLY
SYST
EMS
SY
ST
EM
I(a)
SY
ST
EM
I(b
)S
YS
TE
MI(
c)
SY
ST
EM
I(d
)S
ELE
CT
ION
AL
LELECT~IC
CR
ITE
RIO
NA
LL
-EL
EC
TR
ICA
LL
-EL
EC
TR
ICF
OS
SiL
HE
AT
ING
BA
TT
ER
YF
EE
DB
AC
KF
EE
DB
AC
KP
OW
ER
TR
AC
KIN
GB
AT
TE
RY
EC
ON
OM
iC/P
ER
FO
RM
AN
CE
FA
IRG
OO
DF
AIR
FA
IR-
PO
OR
RE
SU
LTS
CU
RR
EN
TT
eCH
NO
LOG
YF
AIR
FA
IRF
AIR
FA
IRS
TA
TU
S
PR
OJE
CT
ED
ST
AT
US
FA
IR-
GO
OD
GO
OD
GO
OD
FA
IRFO
R19
86
DE
VE
LOP
ME
NT
RIS
KF
AIR
-G
OO
DG
OO
DG
OO
DF
AIR
SY
ST
EM
CO
MP
LE
XIT
YF
AIR
GO
OD
GO
OD
PO
OR
PR
OJE
CT
ED
SY
ST
EM
CO
ST
FA
IRG
OO
DG
OO
DF
AIR
RE
GIO
NA
PP
LIC
AB
ILIT
YG
OO
DG
OO
DF
AIR
-G
OO
DG
OO
D
DU
PL
ICA
TIO
NG
OO
DF
AIR
...G
OO
DF
AIR
FA
IRO
FS
YS
TE
MC
OM
PO
NE
NT
S
RANK
ING
21
34
-,_.-
_-'
0•_
_•.
.._
__
_•••~
•.•
,.
SEPARATE PV/THERMAL SYSTEMS
System II(a) All Electric/Solar Assisted HP/Feedback
Block Diagram
BACK·UP
PV DC/AC100- ..... GENERAL
ARRAY INVERTER LOADS
MAX. POWERTRACKER ~
HEAT~I'-
PUMP
- HOT~ r-
WATER
THERMAL I- THERMALARRAY STORAGE
Description
This system involves all electrically driven loads with solar thermal energyused to supplement space and hot water heating. A heat pump provides thespace heating and cooling requirements. The PV system includes a maximumpower tracking inverter and permits feedback of excess PV energy to the utilitygrid which is connected in parallel with the photovoltaic system. Separatethermal collectors, either liquid or air, provide relatively low temperatureenergy to either a liquid or rock bed storage unit, which in turn provides solarspace heating to supplement the heat pump and preheates domestic hot water. Avariation of this system is one in which the solar collectors provide onlysupplemental energy for domestic hot water heating.
Studied By
• General Electric (Reference #16)
A-9
Advantages and Disadvantages
Advantages
• Utility essentiallyacts as electricalstorage medium
s No fossil back-upsystem required
Conclusions
Disadvantages
• Acceptance offeedback byutil ity
• Thermal dumprequired
This system is economically viable in all regions, approaching the performanceand cost results of the PV only all-electric/feedback system. Since solarthermal energy is used only for domestic hot water during the summer, smallthermal collector areas are required for an optimized economic system. Theseconclusions are based on a solar thermal to PV collector cost ratio of 2 to 1(Reference #16). However, as this ratio is reduced to one or lower, optimizedsolar thermal collector areas increase to 30% or more of the roof area.
System II(b) Direct Solar Heating/Feedback
Block Diagram
BACK-UP
Description
This system involves the general household electrical loads and an electricallydriven vapor compression space cooling unit. The PV system includes a maximumpower tracking inverter and utility feedback capability. Solar thermal collectors,backed-up by a conventional fossil fuel burner, provide space heating and domestichot water requirements.
A-IO
Studied By
• General Electric (Reference #16)
• Aerospace (Reference #22)
Advantages and Disadvantages
Similar considerations as listed for the System IIa, Solar Assisted Heat Pump/Feedback. Additional disadvantages are due to need for fossil back-up system.
Conclusions
This system is less economical than seoarate PV/T systems providing an allelectrical load (IIa), where solar thermal only supplements the heat pump andelectrical hot water units. Separate solar systems used for space cooling andheating (i.e., PV for vapor compression cooling, solar thermal for heating)creates seasonal mismatch of collector areas required. As in the case ofsystem IIa, comparitively small thermal collector area required for optimizedeconomic system. However, increased thermal collector areas become moreeconomic as their cost is reduced and fissil fuels escalate in price.
This system is primarily appllcable in cold and moderate climate areas, wherefossil burners and presently in use and a low to moderate cooling load exists.
System II{c) Solar Rankine Driven HP/Feedback
Block Diagram
BACK·UP
Description
This system ~s.similar to System la, Solar Assisted Heat Pump/Feedbackw~th the add:tlon of a ~eat pump that can be driven electrically or 'wlth appro~rlate clutchlng by a solar driven Rankine engine during thespace coollng season. Highe~ tempe~ature collectors (e.g., vacuum tube)are.used to operate the Ranklne englne at higher efficiency levelsDurlng the heati~g season, solar thermal supplements the conventio~alheat pump operatlon.
A-ll
Studied By'
• General Electric (Reference #16)
Advantages and Disadvantages
Advantages
• Permits use ofhigher temperature,high efficiency,vacuum tube collectorsrequired fo~ Rankinedriven heat pump
• Allows us~ of thermalenergy in summer
Conclusions
Disadvantages
• Complexity of Rankinesystem
• Potentially increasedmaintenance requirements
• Thermal dump systemrequired
Significant cooling requirements must exist in order to achieve economicviability. The regions designated in this program, have moderate to highsummer cooling loads and therefore this system would be applicable. However,maximum return would dictate its application in the warmer regions where bothhigher insolation and higher cooling requirements exist.
From both the performance and cost standpoint, this system compares favorably withSystem IIa, Solar Assisted Heat Pump/Feedback.
System II(d) Solar Absorption Cooling/Feedback
Block Diagram
BACK UP
GENERALLOADS
ABSORPTIONCHILLER
HEATING
HOTWATER
A-12
Description
In this system, the general household electrical loads are provided by thePV array associated with a maximum power tracking inverter. The system also hasfeedback capability. Solar thermal collectors, backed-up by a conventionalfossil fuel burner, provide space heating, hot water, and space cooling usingan absorption chiller.
Studied By
~ General Electric (Reference #16)
Advantages and Disadvantages
Advantages
o Permits use of highertemperature, higherefficiency, vacuum tubecollectors requires forabsorption chiller
$ Allows use of thermalenergy in summer
Conclusions
Disadvantages
I Higher level of technology over solarheating
• Solar cooling technologycurrently not as advancedas heating technology
Comments regarding regional applicability of the Solar Rankine Driven Heat Pump/Feedback System, IIc, apply equally well to this system.
The cooling system employing an absorption chiller yields better system performance than one using a Rankine driven heat pump. This is due to the factthat the absorption chiller requires a lower solar temperature than the Rankinesystem, and thus the collector can operate at hiqher efficiency. However, fromboth the performance and economic standpoint, the Rankine driven heat pumpsystem provides somewhat better overall results.
System II(e) Air Solar Boosted Heat Pump/Battery Storage
A-13
Block Diagram
UTILITY
HEATING
HEATPUMP
COOLING
.... _---,REJECT EXCESS IHEAT INSUMMER IAT NIGHT (FLUSH) I
Description
This system involves all electrically driven loads, with solar thermal energyused to supplement space and hot water heating. A heat pump provides thespace heating and cooling requirements. The PV system is the same as thatdescribed in connection with System la, All Electric System/Battery Storage.Air thermal collectors provide energy to the rock bed storage which in turnserves as a heated air source (i.e., solar boost) for the heat pump duringthe heating season. The solar thermal system can also provide space heatingdirectly when higher temperatures are available. During the cooling season,the rock bed storage serves as a sink for the heat pump during the daytimehours. The nightime hours are used to flush out the heat accumulated in storageduring the day.
Studies By
o Spectro1ab (Reference #20)
Advantages and Disadvantages
Advantages
• Rock bed heat sourcefor heat pump improvesCOP
Disadvantages
o Rejects excess collectedto atmosphere during cooling season
~ Large rock bedt Adequate stratification
difficult to achieveA-14
Conclusions
If the rock bed storage is large and sufficiently stratified, this systemcombines the best features of both the solar boost (also called the series)and solar assisted (also called the parallel or solar supplemented) heatpump systems. However, detailed study (Reference #20) indicated that rockbed size and stratification proved inadequate to achieve the theoreticalpromise of this system.
System II(f) Solar Assisted Heat Pump/Battery Storage
Block Diagram
UTILITY
THEATPUMP
COOLINGHEATING
Description
Similar to System IIa, All Electric/Solar Assisted Heat Pump, with the use ofair thermal collectors and rock .bed storage in place of hydronic collectors andwater storage system. Battery storage, as in System IIc, is also used in thisPV system instead of utility feedback.
Studied By
• Spectrolab (Reference #20)
Advantages and Disadvantages
Advantages
• Lower thermaloperating temperature
; Possible lower equipment costs
Disadvantages
• Large rock bed storage
o DHW only summer demand
A-15
Conclusions
~his s~stem is econom~cally attractive in ~ll regions, with increased viability~n reglons where heatlng loads exceed coollng loads. Use of utility feedbackl~ p1a~e of battery storage would further enhance performance and economicvlablllty of this system.
Since solar thermal used only for domestic hot water during the summer, smallthermal collector areas required for optimized economic system, particularlyin warmer climate regions.
System II(g) Solar Boosted Heat Pump/Feedback
Block Diagram
UTILITY
I PV I I DC/AC I I GENERALARRAY llNVERTER I I LOADSMAX POWER --lHOT WATERI-TRACKER
HEAT
ICOOLING'-- PUMP
TOWER f COOLINGr--- HEATING
I: COLOSTORACE
(WATER SOURCEITHERMAL I I THERMAL I HEAT PUMP)ARRAY I STORAGE
HYDRONIC
Description
This system involves all electric loads, with space heating and cooling providedby a heat pump. The PV system includes a maximum power tracking inverter andutility feedback capability. The hydronic solar thermal system can operate ineither of two modes (i.e., dual modes), (1) as a water source for the heat pumpor (2) provide direct solar space heating. In addition, the solar thermal systemsupplements the electrical resistance heated domestic hot water supply. In thecooling mode, a wet cooling tower and cold thermal energy storage are used toincrease the overall efficiency of the heat pump. Heat is rejected from theheat pump condenser to the ambient air through the coolin9 tower. Cold waterstorage permits shifting operation of the compressor to periods when either solarphotovoltaic power is available or inexpensive off-peak utility power can bepurchased at lower rates.
A-16
Winter operation involves the use of electric resistance heating when thermalenergy storage tank temperature is below 45°F; thermal storage acts as asource for the heat pump between 45°F and 80°F, and direct solar is fed to theair handling duct when thermal storage temperature is above 80°F. For summercooling, the cold water storage tank temperature is maintained between 40°Fand 60°F. If photovo1taic power is available to run the compressor and tank temperature is above 40°F, the heat pump will be operated to remove heat from thetank. If tank temperature goes above 60°F, the compressor is operated independently of the availability of photovo1taic power. When tank temperature goesbelow 40°F, any photovo1taic power available is fed to the utility.
A variation of this system uses a dual mode heat pump, capable of uti1itzingeither ambient air or solar heated water as a source. When thermal storage dropsbelow a minimum temperature level, the heat pump switches to ambient'air source ifoutside temperature level is high enough to provide COP greater than 1, rather thanswitching to auxiliary resistance heating at a COP of 1.
Studied By
• MIT-Lincoln Laboratory and University of Texas at Arlington(Reference #12 and #24)
Advantages and Disadvantages
Advantages
• Higher COP forheat pump
• Cold thermal storagepermits operation ofheat pump to store PVpower
Conclusions
Disadvantages
• Cooling tower to rejectheat from heat pumpcondenser in cooling mode
• Cold water storage tank
s In solar direct heating mode,needs larger than normal heattransfer surface area
t In heating mode,when collectoror storage inadequate, heat pumpmust shut-off, requiring auxiliary electric resistance heating (COP=l)
This system is under test at the University of Texas at Arlington Solar EnergyResearch Fac il ity as part of MIT -L inco 1n Laboratories "Solar Hybrid EnergyProject."
A-17
This system is primarily applicable in areas where both substantial heatingand cooling loads exist. To take maximum advantage of the solar boost mode,the area should experience a significant amount of days in which the temperaturedrops below the 20°F mark.
Comparison Evaluation of PV Only Systems
A comparitive evaluation of each of the separate PV/Thermal systems against aset of criteria previously described in Section 2.0 is presented in Table A-2.The evaluation resulted in the following ranking of separate PV/Thermal systems.
Rank System Title
1 IIa, All Electric/Solar Assisted HeatPump/Feedback
2 lId, Solar Absorption Cooling/Feedback
3 IIc, Solar Rankine Driven Heat Pump/Feedback
4 lIb, Direct Solar Heating/Feedback
5 IIf, Air Solar Assisted Heat Pump/BatteryStorage
6 IIg, Solar Boosted Heat Pump/Feedback
7 lIe, Air Solar B-osted Heat Pump/Battery
A-18
~p
TABL
EA
-2.
COM
PARI
SON
EVAL
UATI
ONS
OFSI
DE-
BY-S
IDE
SYST
EMS
> I ..... to
SYST
EM
IICa
)SY
STE
M1I
1b)
SYST
EM
\IIe
)SY
STE
MII
ld)
SYST
EM
IIle
iSY
STE
MII
If)
SYST
EM
Il(g
)
SEL
EC
TIO
NA
LLE
LE
CT
RIC
DIR
EC
TSO
LA
RSO
LAR
RANK
INE
AB
SOR
PTIO
NA
IRSO
LA
RA
IRSO
LA
RSO
LA
RB
OO
STE
DC
RIT
ER
ION
SOL
AR
ASS
IST
ED
HE
AT
INC
DR
IVE
NH
PC
OO
LIN
GB
OO
STE
DH
EA
TA
SSIS
TE
DH
EA
TH
EAT
PUM
PH
PFE
ED
BA
CK
FEE
DB
AC
KFE
ED
BA
CK
FEE
DB
AC
KPU
MP
IBA
TT
ER
YPU
MP/
BA
TT
ER
YFE
EDBA
CK_
EC
ON
OM
ICIP
ER
FOR
MA
NC
EC
OO
DFA
IRC
OO
DFA
IR-
CO
OD
FAIR
-PO
OR
FAIR
1R
ESU
LT
S
CU
RR
EN
TT
EC
HN
OL
OC
YFA
IRFA
IRFA
IRFA
IRFA
IR-
POO
RFA
IRFA
IRS
TA
TU
S
PRO
JEC
TE
DST
AT
US
GO
OD
GO
OD
GO
OD
CO
OD
FAIR
FAIR
-C
OO
DG
OO
DFO
R19
86
DE
VE
LO
PME
NT
RIS
KG
OO
DG
OO
DFA
IRFA
IR-
GO
OD
FAIR
-PO
OR
CO
OD
FAIR
SYST
EM
CO
MPL
EX
ITY
CO
OD
CO
OD
FAIR
FAIR
POO
RfA
IRFA
IR
PR
OJE
CT
ED
SYST
EM
CO
STC
OO
DG
OO
DFA
IRFA
IRFA
IRC
OO
DFA
IR
RE
CIO
I'!A
PPL
ICA
BIL
ITY
CO
OD
FAIR
GO
OD
CO
OD
FAIR
FAIR
FAIR
MIN
IMU
MD
UPL
ICA
TIO
NO
FG
OO
DFA
IRC
OO
DG
OO
DG
OO
DG
OO
DG
OO
D
SYST
EM
CO
MPO
NE
NT
S
RANK
ING
14
32
75
6
COMBINED PV/THERMAL SYSTEMS
System III(a) Solar Assisted Heat Pump/Feedback
Block Diagram
UTILITY
HYD
PV I DC/AC I I GENERALARRAY !INVERTER I I LOADS
THERMAL MI\~.90IAI"ARRAY
RONICitt"'C.~l!~
- HEAT-PUMP
COOLINGHEATING f--r---
~ HOT ~ I--WATER
I THERMAL II srORAGE I
Description
This is the same system as IIa, with combined PV/Thermal collectors replacingseparate collectors. A variation of this system involves elimination of themaximum power tracker and utility feedback capability; instead, using batterystorage, a shunt regulator and battery charge controller.
Studied By
• General Electric (Reference 16)
~ Westinghouse (References 18 and 19)
• Spectro1ab (Reference 20)
; MIT-Lincoln Laboratory (Reference 13)
A-20
Advantages and Disadvantages
Advantages
• Potentially lowercost collectors thanequivalent separatecollectors
• Lower installationcosts
• Lower roof spacerequirements
Conclusions
Disadvantages
• Combined collectordegrades performanceof both PV and thermal
• Limits flexibility betweenPV and thermal collectorarea required for the variousregions being considered
• DHW only thermal load insummer
• No acceptable combined collector desi gn presentlyavailable
This system is most applicable in regions with high heating; however, theseregions usually have relatively lower annual insolation. The lower insolationlevels coupled with degraded performance associated with combined PV/Thermalcollectors necessitates large collector areas to satisfy a reasonable solarthermal contribution. During the summer months, collector thermal output isused only to provide domestic hot water with resultant waste of excess solarthermal energy.
Though less viable than PV only or separate collector systems, this system,serving the all-electric load with a solar assisted heat pump, is more costeffective than the combined collector, direct solar heatino/battery storaqesystem (IIIb) whieh uses a fossil fuel b~ner for thermal back-up, or thesolar boosted heat pump system (rrre).
System III (b) Direct Solar Heating/Beedback
Block Diagram
PV DclAC GENERALARRAY INVERTER LOADS
THERMAL MAX. POWERARRAY TRACKER COOLING"---
YAP. COMP
...- HEATING
THERMAL-- - FURNACE f--STORAGE
t - HOTWATER A-21
FOSSIL FUEL
Description
This is the same system as lIb, with combined PV/Thermal collectors replacingseparate collectors. A variation of this system involves elimination of themaximum power tracker and utility feedback capability; instead, using batterystorage, a shunt regulator and battery charge controller.
Studied By
I General Electric (Reference 16)
• Westinghouse (Reference )8)
• MIT-Lincoln Laboratory (Reference 13)
Advantages and Disadvantages
Same as System IlIa, Solar Assisted Heat Pump/Feedback. Additional disadvantagedue to need for fossil back-up system.
Conclusions
This system is considerably less effective than System IlIa, Solar Assisted HeatPump in most locations - except for cold areas,where thermal demands are high.As pointed out in the conclusions associated with System IlIa, combined collectorsare not as effective as the other collector options reviewed in this report.Therefore, this system would seem to offer less promise in the future thanSystem II Ia.
System III (c) Solar Boosted HP/Feedback
Block Diagram
UTILITY
HYO
PV I OC/AC I I GENERALARRAY IINVERTERI LOADSTHERMAL
MAXPQWERARRAYRONIC TRACKER
HEAT
ICOOLING IPUMP
TOWER COOLINGr-- HEATING
I HOT IWATER I
(WATER SOURCEI THERMAL I HEAT PUMP)I STORAGE
A-22
Description
Again this system is the same as System IIg, with combined PV/Thermal collectorsreplacing separate collectors and without a cold water storage tank. A variationof this system involves elimination of maximum power tracker and utility feedbackcapability; instead, using battery storage, a shunt regulator and battery chargecontroller.
An additional variation of this system uses a dual mode heat pump, capable ofutilizing either ambient air or solar heated water as a source. When thermalstorage drops below a minimum temperature level, heat pump switches to ambientair source if outside temperature is high enough to provide COP greater than 1,rather than switching to auxiliary resistance heating at a COP of 1.
Studied By
~ Spectrolab (Reference 20)
~ Westinghouse (Reference 18)
~ MIT-Lincoln Laboratory (Reference 13)
Advantages and Disadvantages
Advantages
~ Water source improvesCOP of heat pump
Conclusions
Disadvantages
• In heating mode whencollector or storageinadequate, heat pumpmust shut-off, requiringauxiliary electric resistance heating (COP=l)
• Cooling tower to reject heatfrom heat pump condenserin cooling mode
• In solar direct heating mode,needs larger than normalheat transfer surface
This system is not as effective as the combined collector solar assistea heatpump, system IlIa, from both the performance and economic s~andpoin~. Thissystem is primarily applicable in areas where both substant1al heat1ng andcooling loads exist, and optimum array sizes are comparatively larg~. Whensmall collector areas are used, insufficient stored thermal energy 1Savailable to the heat pump evaporator to allow full-time solar boost operation.Under this condition, the space heating load is met by electric resi~tanceheating at a COP of 1. The large size thermal energy storage.establ1shed asoptional to reduce auxiliary resistance heating adds substant1allv to the costof the solar boost system.
A~3
This system is also considerably more complex and has a higher initial costthan either combined PV/Thermal Systems IlIa or IIIb.
System III(d) Direct Heating/Battery Storage/Stand Alone
Block Diagram
PVARRAY
THERMALARRAY
HYDRONIC
SMALL 1':'1.5 KW)AUXiLIARY
ENGINEGENERATOR
FOSSILFUEL
Description
This is a completely autonomous system with no connection to the utility grid.Combined PV/Thermal collectors are used in conjunction with electrical andthermal storage. Back-up is handled by a small auxiliary diesel electricgenerator (on the order of 1.5 KW) and a fossil fuel burner. Loads are allelectric with the exception of space heating. The PV array (or battery storage)provides electrical energy to the general household loads, a vapor compressioncooling unit, and the domestic hot water heater. Thermal energy generated inthe combined array (or by the fossil burner) provides direct space heating andsupplemental hot water heating.
Both the electrical and thermal storaqe are essential to a stand-alone system.During the day, energy derived from the solar system is stored to get throughthe night. If storage energy is inadequate, the auxiliary supply either pro-vides energy directly or increases the energy store. This allows the auxiliarysystems to be of low capacity. Thus, the storage systems not only provide storage,but handle peak loads.
A-24
The solar system should be capable of providing on the order of 80% to90% of the household electrical and thermal loads. The auxiliary electricalgenerator should run no more than 1000 hours per year and fossil fuel usesapproximately 10% for electrical back-up and 20% for thermal back-up.
In addition to the combined PV/Thermal collectors depicted in the aboveschematic, variations of the system involve the use of PV only collectors,and PV only side by side with combined collectors sized appropriately forthe specific locations.
Another variation of the stand-alone system involves the all-electric residence,pv only array, and heat pump for providing space heating and cooling. Energyfor hot water is partially supplied by excess PV electrical energy in spring andfall and by reject heat from the air conditioning in the summer.
Waste heat from the diesel generator represents another potential source ofthermal energy that could be applied in satisfying residential demand.
Studied By
, Westinghouse (References 18 and 19)
Advantages and Disadvantages
Advantages
• No utility interface
• Potential for lowerenergy cost
Disadvantages
I Potential noise andenvironmental problems
• Must assure continuousinternal energy availability
; Residence designed to minimize exceptional loads
• Load management unit may benecessary to minimize extremesdue to the system1s limitedpeak power
; Extensive amount of equipment and maintenance
A-25
Conclusions
Stand~alone systems are viable virtually everywhere that ut'ility back-upsystems are viable. Stand-alone systems favor combined PV/Thermal collectorsin cold regions. Large collector array areas are required to meet 90% oftotal annual load requirements in colder/lower insolation areas, such asCleveland. In regions with moderate heating requirements, the collector arearequired for optimum electrical input can readily exceed the optimum combinedcollector area needed for direct thermal conversion. Since combined collectorsare more costly, a combination of PV only and combined collectors are moreappropriate for these regions. In very warm areas, where air conditioning loadspredominate, PV only collectors meet requirements.
Comparison Evaluation of Combined PV/Thermal Systems
A comparitive evaluation of each of the combined PV/Thermal systems againsta set of criteria previously described in Section 2.0 is presented in Table A-3.This evaluation resulted in the following ranking of combined PV/Thermal systems.
Rank System Title
1 II la, Solar Assisted Heat Pump/Feedback
2 IIIb, Direct Solar Heating/Feedback
3 II Id, Stand-Alone/Direct Heating/Battery
4 II Ic, Solar Boosted Heat Pump/Feedback
A-26
> II:
\:)
-:l
TABL
EA
-3.
COM
PARI
SON
EVAL
UATI
ONOF
COM
BINE
DSY
STEM
S
SY
ST
EM
III(
a)
SY
ST
EM
III
(b)
SY
ST
EM
III(
c)
SY
ST
EM
III(
d)
SE
LE
CT
iON
SO
LAR
AS
SIS
TE
DD
IRE
CT
SO
LAR
SO
LA
RB
OO
ST
ED
ST
AN
DA
LO
NE
CR
ITE
RIA
HE
AT
ING
HE
AT
PU
MP
ID
IRE
CT
HE
AT
ING
IH
P/FE
ED'B
ACK
FE
ED
BA
CK
FEED
BACK
BA
TT
ER
Y
EC
ON
OM
IC/P
ER
FO
RM
AN
CE
FA
IRF
AIR
-P
OO
RF
AIR
FA
IRR
ES
UL
TS
CU
RR
EN
TT
EC
HN
OL
OG
YP
OO
RP
OO
RP
OO
RP
OO
RS
TA
TU
S
PR
OJE
CT
ED
ST
AT
US
FA
IRF
AIR
FA
IRF
AIR
FO
R19
86
DE
VE
LOP
ME
NT
RiS
KF
AIR
FA
IRF
AIR
FA
IR
SY
ST
EM
CO
MP
LE
XIT
YG
OO
DG
OO
DF
AIR
FA
IR-
PO
OR
PR
OJE
CT
ED
SY
ST
EM
CO
ST
GO
OD
GO
OD
FA
IRF
AIR
RE
GIO
NA
PP
LIC
AB
ILIT
YF
AIR
FA
IRF
AIR
GO
OD
MIN
IMU
MD
UP
LIC
AT
ION
OF
GO
OD
GO
OD
GO
OD
GO
OD
SY
ST
EM
CO
MP
ON
EN
TS
RANK
ING
12
Ll3
.;.I;,