Issue 1 September 2012

OFF GRID PV POWER SYSTEMS

SYSTEM DESIGN GUIDELINES


















These guidelines have been developed by the Sustainable Energy Industry Association of the Pacific Islands in
Collaboration with the Pacific Power Association
They represent latest industry BEST PRACTICE for the design and installation of PV Grid Connect Systems.
Copyright 2012

While all care has been taken to ensure this guideline is free from omission and error, no responsibility can be taken
for the use of this information in the design of any PV grid connect system.

Issue 1 September 2012 Page 1


GENERAL
The design of any off-grid system should consider, other than the electrical load, a number of criteria such as:
o Budget
o Power quality
o Environmental impact
o Aesthetics
o Acceptable genset runtime
o Noise levels
o Site accessibility
o Level of automation
Note: This guidelines are based on d.c. bus systems and do not include the new a.c. bus hybrid systems currently available. Guidelines dedicated to
hybrid Systems will be developed.

ENERGY SOURCE MATCHING
Heating and lighting should be supplied from the most appropriate source. For example -
o cooking - gas or wood burning stove
o water heating - solar water heating with gas or wood backup
o Lighting - electrical lighting most often used but natural light ( daylighting ) should be considered.

ENERGY EFFICIENCY
All appliances should be chosen for the lowest possible energy consumption for each desired outcome, such
as
o High efficiency lighting
o Energy efficient refrigeration

STANDARDS for DESIGN
System designs should follow any standards that are typically applied in the country or region where the solar
installation will occur. The following lists the relevant standards in Australia, New Zealand and USA They are
listed because some Pacific island countries and territories do follow those standards. These standards are
often updated and amended so the latest version should always be applied.
In Australia and New Zealand the main standards required include:
o AS/NZS3000 Wiring Rules
o AS/NZS4509 Stand-alone power systems
o AS 4086.2 Secondary batteries for stand-alone power supplies
o AS/NZS5033 PV Array
o AS 3010.1 Electrical Installations - Supply Generating set
o AS 1768 Lightning Protection
o AS 3595 Energy management programs
o AS 1359.51 Noise level limits

In USA PV systems must be in accordance with the following codes and standards:
o Electrical Codes-National Electrical Code Article 690:Solar Photovoltaic Systems and NFPA 70
Uniform Solar Energy Code
o Building Codes- ICC, ASCE 7
o UL Standard 1701: Flat Plat Photovoltaic Modules and Panels
o UL Standard 1741: Standard for Inverter, converters, Controllers and Interconnection System
Equipment for use with Distributed Energy Resources















Issue 1 September 2012 Page 2

INTRODUCTION

Four major issues arise when designing a system:

1. the load (power) required to be supplied by the system is not constant over the period of one day;
2. the daily energy usage varies over the year;
3. the energy available from the PV array may vary from time to time during the day;
4. the energy available from the PV array will vary from day to day during the year.

Since the system is based on photovoltaic modules, then a comparison should be undertaken between the
available energy from the sun and the actual energy demands The worst month is when the ratio between
solar energy available and energy demand is smallest.

The design of an off-grid power requires a number of steps. A basic design method follows:

1. Determination of the energy usage that the system must supply.
2. Determination of the battery storage required.
3. Determination of the energy input required from the PV array or other sources (eg battery
charger/generator)
4. Selection of the remainder of system components.

LOAD (ENERGY) ASSESSMENT

Electrical power is supplied from the batteries (DC) or via an inverter to produce either 230 volts AC (South
Pacific) or 110 / 120 volts AC (North Pacific). Electrical energy usage is normally expressed in watt hours (Wh)
or kilowatt hours ( kWh ).

To determine the daily energy usage for an appliance, multiply the power of the appliance by the number of
hours per day it will operate. The result is the energy (Wh) consumed by that appliance per day.

Appliances can either be DC or AC. An energy assessment should be undertaken for each type, examples of
these are shown in tables 1 and 2.

You need to calculate the electrical energy usage with the customer. Many systems have failed over the years
not because the equipment has failed or the system was installed incorrectly, BUT BECAUSE THE
CUSTOMER BELIEVED THEY COULD GET MORE ENERGY FROM THEIR SYSTEM THAN THE SYSTEM
COULD DELIVER. It failed because the customer was unaware of the power/energy limitations of the
system.

The problem is that the customer may not want to spend the time determining their realistic power and energy
needs which is required to successfully complete a load assessment form. They just want to know: How much
for a system to power my lights and TV?

A system designer can only design a system to meet the power and energy needs of the customer. The
system designer must therefore use this process to understand the needs of the customer and at the same
time educate the customer. Completing a load assessment form correctly (Refer to table 1 and 2 below) does
take time; you may need to spend 1 to 2 hours or more with the potential customer completing the tables. It is
during this process that you will discuss all the potential sources of energy that can meet their energy needs
and you can educate the customer on energy efficiency.

Issue 1 September 2012 Page 3


Worked Example
Table 1 DC Load (energy) Assessment


Table 2 AC Load (energy) Assessment
(1) (2) (3) (4a) (5a) (4b) (5b) (6) Comments


Appliance


Number

Power
dry season wet season Contribution
to maximum
demand

Usage
Time

Energy

Usage
Time

Energy
W h Wh h Wh W
Light 4 7 4 112 4 112 28
Daily Load energy-d.c loads (Wh) ( DC 7a) 112 (DC 7b) 112
Maximum d.c. demand (W) (DC 8) 28
(1) (2) (3) (4a) (5a) (4b) (5b) (6) (7) (8) (9a) (9b)



Comments


Appliance


No.

Power
dry season wet season

Power
Factor
Contribution
to max
demand


Surge
Factor
Contribution to
surge demand

Usage
Time

Energy

Usage
Time

Energy
Potential Design
W h Wh h Wh VA VA VA
TV 100 3 300 3 300 0.8 125 4 500 125
Refrigerator 100 12 1200 12 1200 0.8 125 4 500 500 Duty cycle of
0.5 included
Daily Load Energy A.C
Loads (Wh)

(AC10a)
1500
(AC10b)
1500
maximum demand (VA) (AC11) 250 1000
Surge demand (VA) (AC12) 625

Issue 1 September 2012 Page 4


In the worked example on the previous page, the TV and refrigerator are using AC electricity so we have to
take into account the efficiency of the inverter. Typically the peak efficiency of the inverter may be over 90%
but in many systems the inverter will sometimes be running when there is very little load on the inverter, so the
average efficiency is about 85% to 90%. Then we must divide the total AC energy used by this figure to obtain
the energy required to be supplied to the inverter from the battery bank.

For the worked example assume the efficiency of the chosen inverter is 90%.

Daily battery load (energy) from AC loads =
1500Wh 0.9 = 1667 Wh

Daily battery load (energy) from DC loads = 112 Wh


To get the total load(energy) as seen by the battery, you add the two figures together:
1667 + 112= 1779Wh
If there are no AC loads, then you only have to work out the load from the DC appliances, and not include the
inverter (or the inverter efficiency).

BATTERY SELECTION

DETERMINATION OF SYSTEM VOLTAGE

System voltages are generally 12, 24 or 48 Volts. The actual voltage is determined by the requirements of the
system. For example, if the batteries and the inverter are a long way from the energy source then a higher
voltage may be required to minimise power loss in the cables. In larger systems 120V or 240V DC could be
used, but these are not typical household systems.

As a general rule, the recommended system voltage increases as the total load increases. For small daily
loads, a 12V system voltage can be used. For intermediate daily loads, 24V is used and for larger loads 48V is
used.









The changes over points are roughly at daily loads of 1 kWh and 3-4 kWh but this will also be dependent on
the actual power profile.

One of the general limitations is that maximum continuous current being drawn from the battery should not be
greater than 150A.

BATTERY SIZING
To convert Watt-hours (Wh) to Amp-hours (Ah) you need to divide by the battery system voltage.

For the worked example the daily energy usage was 1779Wh , so we select a battery system voltage of 24
Volts.

This means that the daily Ah demand on the batteries will be:

1 kWh
3-4 kWh
Use 12 Volt
system voltage
Use 24 Volt
system voltage
Use 48 Volt
system voltage

Issue 1 September 2012 Page 5

Ah =
Wh system voltage 1779Wh 24 = 74 Ah

So at the moment the minimum size battery to meet the daily energy requirements in the example is: 74Ah

Battery capacity is determined by whichever is the greater of the following two requirements:

1. The ability of the battery to meet the energy demand of the system, often for a few days, sometimes
specified as days of autonomy of the system;
OR
2. The ability of the battery to supply peak power demand.

The critical design parameters include:

Parameters relating to the energy requirements of the battery:
a) Daily energy demand
b) Daily and maximum depth of discharge
c) Number of days of autonomy

Parameters relating to the discharge power (current) of the battery:
a) Maximum power demand
b) Surge demand

Parameters relating to the charging of the battery:
a) Maximum Charging Current

Based on the these parameters there are a number of factors that will increase the battery capacity in order to
provide satisfactory performance. These correction factors must be considered.


1. Days of Autonomy
Extra capacity is necessary where the loads require power during periods of reduced input. The battery
bank is often sized to provide for a number of days autonomy. A common period selected is 5 days.

Where a generator is operating on a regular basis the autonomous period can be reduced.

In other cases, where there is no auxiliary charging source, the period of autonomy is often increased to 7
days or more.

For the worked example assume 5 days autonomy .

ADJUSTED

Battery Capacity = 74 x 5 = 370 Ah



2. Maximum Depth of Discharge
Battery manufacturers recommend a maximum depth of discharge (DOD). If this is regularly exceeded the
life of the battery is severely reduced. This could be 50% or for some solar batteries as high as 80%.

For the worked example Assume a maximum DOD of 70%.

ADJUSTED
Battery Capacity = 370 0.7 = 529 Ah


3. Battery Discharge Rate

The actual discharge rate selected is highly dependent on the power usage rates of connected loads. Many
appliances operate for short periods only, drawing power for minutes rather than hours. This affects the battery

Issue 1 September 2012 Page 6

selected, as battery capacity varies with discharge rate. Information such as a power usage profile over the
course of an average day is required for an estimate of the appropriate discharge rate.

For small systems this is often impractical.

Where the average rates of power usage are low, the battery capacity for 5 days autonomy is often selected at
the 100hr rate of discharge for the battery.

For the worked example
ADJUSTED

Battery Capacity = 529 Ah (@ C
100
)

Where average power usage rates are high, it may be necessary to select the battery capacity for 5 days
autonomy at a higher discharge rate. eg. the 10 (C
10
) or 20hr (C
20
) rate

4. Battery Temperature derating
Battery capacity is affected by temperature. As the temperature goes down, the battery capacity reduces. The
following graph gives a battery correction factor for low temperature operation. Note that the temperature
correction factor is 1 at 25C as this is the tempe rature at which battery capacity is specified.


In the tropics it is often still 20C+ in the even ings so unless the system is located in a mountainous region
that does get cold then ignore the temperature derating. If you want to be conservative add 5% to the capacity
to allow for this factor.



BATTERY SELECTION
Deep discharge type batteries / cells should be selected for the required system voltage and capacity in a
single series string of battery cells.

Parallel strings of batteries are not recommended.

Where this is necessary each string must be separately fused.

For the worked example a battery of at least 529 Ah (@C
100
)

should be used.


PV ARRAY SIZING- Standard Switched Controllers

The calculation for determining the size of the PV array is dependent on the type of controller used. Historically
standard switched controllers were the most common controllers used. In recent years a number of maximum

Issue 1 September 2012 Page 7

power point trackers (MPPT) have become available. This section determines how to size the PV array based
on switched controllers based on the PV array meeting the daily load requirements all year. Later in the guide
is a section on how to size a PV array using a MPPT.

The size of the PV array should be selected to take account of:
seasonal variation of solar irradiation
seasonal variation of the daily energy usage
battery efficiency
manufacturing tolerance of modules
dirt
temperature of array (the effective cell temperature)

Solar irradiation data is available from various sources. Some countries have data available from their
respective meteorological department. One source for solar irradiation data is the NASA website:
http:/eosweb.larc.nasa.gov/sse/. RETSCREEN, a program available from Canada, incorporates the NASA
data and it is easier to use. Please note that the NASA data has, in some instances, had higher irradiation
figures than that recorded by ground collection data in some countries. but if there is no other data available it
is data that can be used.

Solar irradiation is typically provided as kWh/m
2
however it can be stated as daily peak Sun Hours (PSH). This
is the equivalent number of hours of solar irradiance of 1kW/m
2
.

Attachment 1 provides data on the following sites:
Suva, Fiji (Latitude 1808 S Longitude 17825 E)
Apia, Samoa (Latitude 13
o
50' S' Longitude 171
o
44' W)
Port Vila, Vanuatu (Latitude 1744' S Longitude 16819' E)
Tarawa, Kiribati (Latitude 128'N, Longitude 1732 'E)
Raratonga, Cook islands( Latitude 2130'S, Longitude 1600'W)
Nukualofa, Tonga (Latitude 2114'S Longitude 17522'W)
Honiara, Solomon Islands (Latitude 0927'S, Lon gitude 15957'E)
Koror ,Palau ( Latitude 720N Longitude 13428'E)
Ponapei, Pohnpei FSM (Latitude: 654'N, Longitude : 15813'E)
Majuro, Marshall Islands (Latitude: 7 12N, Longitude 171 06E)
Alofi, Niue (Latitude 1904' S. Longitude 16955' W)
Nauru (Latitude 055S, Longitude 166 91E)
Tuvalu (Latitude 831 S, Longitude 17913 E)
Hagta, Guam (Latitude 1328 N Longitude: 14445 E)
Noumea, New Caledonia (Latitude 2216 S Longitude: 16627 E)
Pago Pago, American Samoa (Latitude 1416 S Longitude: 17042 W)

The variation of both the solar irradiation and the load energy requirement should be considered. If there is no
variation in daily load between the various times of the year then the system should be designed on the month
with the lowest irradiation that is peak sun hours (PSH).

DAILY ENERGY REQUIREMENT FROM THE PV ARRAY
In order to determine the energy required from the PV array, it is necessary to increase the energy from the
battery bank to account for battery efficiency.
.
The average columbic efficiency (in terms of Ah) of a new battery is 90% (variations in battery voltage are not
considered).

For the worked example the daily energy requirement expressed in Ah from the battery is 74 Ah. Allowing for
the battery efficiency, the solar array then needs to produce
74 Ah 0.9 = 82.2 Ah

Assume the worst months PSH is 5.


Issue 1 September 2012 Page 8

Therefore the required PV array output current is:

82.2 Ah 5 PSH = 16.5 A

OVERSIZE FACTOR
If the system does not include a fuel generator which can provide extra charging to the battery bank then the
solar array should be oversized to provide the equalisation charging of the battery bank. In Australia and New
Zealand this is between 30% and 100%. It is recommended in the Pacific that this is 10%.

For the worked example the adjusted array output current is:

16.5A x 1.1 = 18.1 A

DERATING MODULE PERFORMANCE
The PV array will be de-rated due to:
Manufacturers Tolerance: Most manufacturers rate their modules a percentage (eg 3%) or wattage
(eg 2W). Unless every module is tested and its actual rating is known then the modules should be de-
rated by the manufacturers tolerance.
Dirt: Over a period of time dirt or salt (if located near the coast) can build up on the array and reduce
the output. The output of the module should therefore be derated to reflect this soiling. The actual value
will be dependent on the site but this can vary from 0.9 to 1 (i.e. up to 10% loss due to dirt).
Temperature: Modules output power decreases with temperature above 25C and increases with
temperatures below 25C. The average cell temperatu re will be higher than the ambient because of the
glass on the front of the module and the fact that the module absorbs some heat from the sun. The
output power and/or current of the module must be based on the effective temperature of the cell. This
is determined by the following formula:

T
cell-eff
= T
a.day
+ 25C

Where
T
cell-eff
= the average daily effective cell temperature in degrees Celsius (C)
T
a.day
= the daytime average ambient temperature for the month that the sizing is being undertaken.

Since the modules are used for battery charging, the current at 14 Volts (a good battery charging voltage) at
the effective cell temperature should be used in calculations. If curves are unavailable to determine the current
at effective cell temperature then use the Normal Operating Cell temperature (NOCT) provided by the
manufacturers.

Therefore the derated module output current is calculated as follows:

The Current of the module at 14V and effective cell temperature (or NOCT current)
multiplied by derating due to manufacturers tolerance
multiplied by derating due to dirt

I
(NOCT)
x f
man
x f
dirt

If a module has a 3% (0.03) manufacturers tolerance, then the module current is derated by multiplying by
0.97 (1-0.03).

If a module has a 5% (0.05) loss due to dirt then the module current is derated by multiplying by 0.95 (1-0.05).

For the worked example the selected module has a peak rating of 80W
P
.

An 80 watt solar module (still used in many small off-grid systems) typically has the following data:





Issue 1 September 2012 Page 9

Table 3: 80 W module data
Rated Power 80W
Power Tolerance 5%
Nominal Voltage 12V
Maximum Power Voltage, V
mp
17.6V
Maximum Power Current, I
mp
4.55A
Open Circuit Voltage, V
oc
22.1V
Short Circuit Current, I
sc
4.8A
NOCT 472C
Current at 14V and NOCT 4.75A


Assuming a 5% dirt derating then the adjusted output current of the above module is:

ADJUSTED

Module current = I
(NOCT)
x 0.95 ( minus 5% for manufactures tol) x 0.95 ( minus 5% for dirt)

= 4.75A x 0.95 x 0.95 = 4.29A

NUMBER OF MODULES REQUIRED IN ARRAY

First determine number of modules in series, To do this divide the system voltage by the nominal operating
voltage of each module. In our example:

For the worked example Number of modules in series = 24V 12V = 2

Therefore the array must comprise of series connected strings of 2 modules.

To determine the number of strings in parallel, the PV array output current required (in A) is divided by the
output of each module (in A).

For the worked example Number of strings in parallel =18.1A 4.29A = 4.22

Do we round up or down? If you want to be conservative you would round up. However in this example we
suggest you round down since this calculation was based on the worst month and we allowed an oversize of
10%.This does need to be determined for each system.

For the worked example The number of modules in the array = 4 x 2 = 8

The peak rating of the array is : 8 x 80W
P
= 640W
P



INVERTER SELECTION

The type of inverter selected for the installation depends on factors such as cost, surge requirements, power
quality and for inverter/chargers, a reduction of the number of system components necessary. Inverters are
available in 3 basic output types - Square wave, modified square wave and sine wave.
There are few square wave inverters used today.

Modified square wave inverters generally have good surge and continuous capability and are usually cheaper
than sine wave types. However, some appliances, such as audio equipment, television and fans can suffer
because of the output wave shape.

Sine wave inverters often provide a better quality power than the 230V (or 110V or 120V) grid supply.


Issue 1 September 2012 Page 10

INVERTER SIZING
The selected inverter should be capable of supplying continuous power to all AC loads
AND
providing sufficient surge capability to start any loads that may surge when turned on and particularly if they
turn on at the same time.

Where an inverter cannot meet the above requirements attention needs to be given to load control and
prioritisation strategies.

For the worked example
From the load (energy) assessment on page 3, a selected inverter must be capable of supplying 250VA
continuous with a surge capability of 625VA.


CONTROLLERS- Standard Switched Controller

PV controllers on the market range from simple switched units that only prevent the overcharge (and
discharge) of connected batteries to microprocessor based units that incorporate many additional features
such as

o PWM and equalisation charge modes
o DC Load control
o Voltage and current metering
o Amp-hour logging
o Generator start/stop control

Unless the controller is a model that is currently limited these should be sized so that they are capable of
carrying 125% of the array short circuit current and withstanding the open circuit voltage of the array. If there is
a possibility that the array could be increased in the future then the controller should be oversized to cater for
the future growth.

(Note: sometimes the controller is called a regulator)

For the worked example
The controller chosen must have a current rating
> 1.25 x 4 x 4.8 A = 24A at a system voltage of 24V.



GENERATORS & BATTERY CHARGING

To reduce system costs, it is common for some form of auxiliary charging to be used to provide energy when
daily energy requirements are greater than the daily PV input into the system. This is usually a
diesel/petrol/gas powered generator. Where the electrical output is 230V AC (or 110V or 120V) a battery
charger is required.

An inverter/charger can be used; otherwise a separate charger unit is needed.

Factors that must be considered when using internal combustion generators are
o Fuel storage and spillage precautions
o Noise emission control
o Ventilation
o Generator loading

With regards to generator loading, a generator should supply greater than 50% of its maximum rating while
running. Loading of less than 50% increases running and maintenance costs and reduces generator life.

Issue 1 September 2012 Page 11

( refer to genset manufacturers' information )

BATTERY CHARGER SIZING
A charger must be capable of supplying voltage greater than the nominal system voltage.
The maximum charging current must not be greater than that recommended by the battery manufacturer but a
usable estimate is a maximum charge current of around 10% of the C
10
rate.


PV ARRAY SIZING- MPPT

DAILY ENERGY REQUIREMENT FROM THE PV ARRAY

The size of the PV array should be selected to take account of:
seasonal variation of solar irradiation
seasonal variation of the daily energy usage
battery efficiency (wh)
Cable losses
MPPT efficiency
manufacturing tolerance of modules
dirt
temperature of array (the effective cell temperature)

With the standard controller the only sub-system losses was the battery efficiency and the calculations are
undertaken using Ah. When using a MPPT the calculations are in Wh and the sub-system losses in the system
include:
Battery efficiency (watthr)
Cable losses
MPPT efficiency

In order to determine the energy required from the PV array, it is necessary to increase the energy from the
battery bank to account for all the sub-system losses.

For the worked example assume cable losses is 3% (transmission efficiency of 97%), MPPT efficiency of 95%
and battery efficiency of 80%

Subsystem efficiency = 0.97 x 0.95 x 0.8 = 0.737

Energy required from the PV array = 1779Wh

0.737 = 2413Wh

Assume the worst months PSH is 5.

Therefore the required peak PV array output power is:

2413Wh 5 PSH = 482W
P


OVERSIZE FACTOR
If the system does not include a fuel generator which can provide extra charging to the battery bank then the
solar array should be oversized to provide the equalisation charging of the battery bank. In Australia and New
Zealand this is between 30% and 100%. It is recommended in the Pacific that this is 10%.

For the worked example Therefore the adjusted array output current is:

482W
P
x 1.1 = 530W
P


DERATING MODULE PERFORMANCE
The PV array will be de-rated due to:

Issue 1 September 2012 Page 12

Manufacturers Tolerance: Most manufacturers rate their modules a percentage (eg 5%) or wattage
(eg 2W). Unless every module is tested and its actual rating is known then the modules should be de-
rated by the manufacturers tolerance.
Dirt: Over a period of time dirt or salt (if located near the coast) can build up on the array and reduce
the output. The output of the module should therefore be derated to reflect this soiling. The actual value
will be dependent on the site but this can vary from 0.9 to 1 (i.e. up to 10% loss due to dirt).
Temperature: Modules output power decreases with temperature above 25C and increases with
temperatures below 25C. The average cell temperatu re will be higher than the ambient temperature
because of the glass on the front of the module and the fact that the module absorbs some heat from
the sun. The output power and/or current of the module must be based on the effective temperature of
the cell. This is determined by the following formula:

T
cell-eff
= T
a.day
+ 25C

Where
T
cell-eff
= the average daily effective cell temperature in degrees Celsius (C)
T
a.day
= the daytime average ambient temperature for the month that the sizing is being undertaken.

With switched controllers the temperature effect was used to determine the operating current of the
module/array. With MPPTs the derating power factor must be calculated.

The three main types solar modules available on the market each have different temperature coefficients.
These are:
Monocrystalline: Modules typically have a temperature coefficient of 0.45%/
o
C. That is for every
degree above 25
o
C the output power is derated by 0.45%.
Polycrystalline: Modules typically have a temperature coefficient of 0.5%/
o
C.
Thin Film: Modules have a different temperature characteristic resulting in a lower co-efficient typically
around 0%/C to -0.25%/C, but remember to check with the manufacturer

The typical ambient daytime temperature in many parts of the Pacific is between 30 and 35
o
C during some
times of the year. So it would not be uncommon to have module cell temperatures of 55
o
C or higher.

For the worked example Assume the ambient temperature is 30
o
C..

Therefore the effective cell temperature is

30
o
C +25
o
C = 55
o
C
Therefore this is 30
o
C above the STC temperature of 25
o
C.

The 80 W
P
module used in this example is a polycrystalline module with a derating of -0.5%/
o
C

Therefore the losses due to temperature would be:

Temperature loss = 30
o
C x 0.5%/
o
C = 15% loss

This is a temperature derating factor of 0.85

Assuming a 5% dirt derating then the adjusted output power of the 80W module is:

Adjusted module power = 80 x 0.95 x 0.95 x 0.85 = 61.4W

NUMBER OF MODULES REQUIRED IN ARRAY

To calculate the required number of modules in the array, divide the required array power by the adjusted
module power.



Issue 1 September 2012 Page 13

For the worked example
Number of modules in array =530 61.4 = 8.63

(Note for the switch regulator we had 4.29 parallel strings of 2 modules in series)

The actual number of modules will be dependent on the MPPT selected. If it was 9 then the rating of the array
is : 9 x 80W
P
= 720W
P



SELECTING MPPT

The following table gives some examples of MPPTs currently available on the market:

Table 4: MPPT Data
Model d.c. battery
Voltage (V)
Input voltage
range (V)
Max d.c.
Battery
Current (A)
Max(W) Solar
Array
Max Load
Current (A)
STECA
Solarix
MMP2010
12/24 17 to 100 20 250/500 10
Phocos
MMPT
100/20-1
12/24 Max 95 20 300/600 10
Morningstar
SS-MPPT-15L
12/24 Max 75 15 200/400 15
Outback Flex
Max 80
12/24/36/48/60 Max 150 80 1250(12)
2550(24)
5000(48)
7500 (60)

Outback Flex
Max60
12/24/36/48/60 Max 150 60 900(12)
1800(24)
3600(48)
4500 (60)


For the worked example
Allowing for 125% oversizing then the required rating of the MPPT is:

1.25 x 720Wp = 900 W

From the table we would select the Outback Flexmax60 which has an array rating of 1800W @ 24V

We could possibly use two of the others e.g. Phocos or Steca.

MATCHING THE PV ARRAY TO THE MAXIMUM VOLTAGE SPECIFICATIONS OF THE MPPT
The MPPT typically have a recommended minimum nominal array voltage and a maximum voltage. In the
case where a maximum input voltage is specified and the array voltage is above the maximum specified, the
MPPT could be damaged.

Some MPPT controllers might allow that the minimum array nominal voltage is that of the battery bank.
However the MPPT will work better when the minimum nominal array voltage is higher than the nominal
voltage of the battery. The Outback range of MPPTs requires that the minimum nominal array voltage is
greater as shown in Table 5. Please check with the MPPT manufacturer because these could vary.





Issue 1 September 2012 Page 14


Table 5 Minimum Nominal Array Voltages (Outback MPPTs)
Nominal Battery Voltage Recommended Minimum Nominal Array
voltage
12V 24V
24V 36V
48V 60V

It is important that the output voltage of the string is matched to the operating voltages of the MPPT and that
the maximum voltage of the MPPT is never reached.

The output voltage of a module is affected by cell temperature changes in a similar way to the output power.
The manufacturers will provide a voltage temperature coefficient. It is generally specified in V/C (or mV/C)
but it can also be expressed as a % .

To ensure that the V
oc
of the array does not reach the maximum allowable voltage of the MPPT the minimum
day time temperatures for that specific site are required.

In early morning at first light the cell temperature will be very similar to the ambient temperature because the
sun has not had time to heat up the module. In the Pacific Islands the average minimum temperature is 20
0
C
(this could be lower in some mountain areas) and it is recommended that this temperature is used to
determine the maximum V
oc.
(Note: If installing in the mountains then use the appropriate minimum
temperature.

Many people also use 0C, if appropriate for the ar ea). The maximum open circuit voltage is
determined similar to the temperature derating factor for the power.

For the worked example Assume the voltage co-efficient is 0.07V/
o
C.

If the minimum temperature is 20
o
C this is 5
o
C below the STC temperature of 25
o
C. Therefore the effective
variation in voltage is:
5 x 0.07 = 0.35V

So the maximum open circuit voltage of the module = 22.1V + 0.35V =22.45V

Maximum number of modules that you can have in series = 150V 22.45V = 6.68 this is rounded down to 6.

So the MPPT will allow between 3 and 6 modules in a string.

The actual number of modules required was 8.68. If we round down to 8 (since an oversize factor of 10% and
also worst month for PSH was used) then the array would be 2 parallel strings of 4 modules in series. If we
round up to 9 then the solution could be 3 parallel strings with 3 modules in series.




Issue 1 September 2012 Page 15

ATTACHMENT 1: : Table showing Peak Sunhrs for various sites and tilt angles.


Location Peak Sunlight Hours (kWh/m/day)
Suva, Fiji Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Annual
Average
Latitude: 1808 South 0Tilt 6.29 6.2 5.54 4.67 4.05 3.72 3.89 4.44 5.08 6.04 6.32 6.38 5.21
Longitude: 17825 East 18Tilt 6.27 5.88 5.55 4.99 4.61 4.38 4.51 4.88 5.22 5.83 6.1 6.41 5.38
33Tilt 5.95 5.4 5.33 5.03 4.85 4.7 4.8 5 5.1 5.43 5.71 6.12 5.29

Apia, Samoa Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Annual
Average
Latitude: 1350 South 0Tilt 5.39 5.47 5.16 5.09 4.63 4.46 4.71 5.25 5.77 5.91 5.76 5.51 5.25
Longitude: 17146 West 13Tilt 5.31 5.24 5.12 5.32 5.07 5 5.24 5.61 5.85 5.72 5.67 5.45 5.38
28Tilt 5.13 4.86 4.93 5.38 5.36 5.42 5.64 5.81 5.75 5.36 5.45 5.3 5.37

Port Vila, Vanuatu Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Annual
Average
Latitude: 1744 South 0Tilt 6.68 6.2 5.76 4.98 4.2 3.79 4.04 4.75 5.65 6.47 6.67 6.93 5.5
Longitude: 16819 East 17Tilt 6.69 5.9 5.78 5.33 4.76 4.42 4.66 5.22 5.82 6.26 6.46 7.01 5.69
32Tilt 6.38 5.43 5.56 5.39 5.02 4.75 4.98 5.39 5.72 5.83 6.07 6.73 5.61

Tarawa, Kiribati Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Annual
Average
Latitude: 0128 North 0Tilt 5.58 5.98 5.99 5.87 5.82 5.7 5.87 6.15 6.52 6.4 6.1 5.5 5.95
Longitude: 17302 East 16Tilt 5.9 6.11 5.83 5.79 5.94 5.92 6.06 6.17 6.28 6.45 6.43 5.88 6.06



Rarotonga, Cook Islands Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Annual
Average
Latitude: 2112 South 0Tilt 6.45 6.14 5.78 4.59 3.86 3.54 3.73 4.46 5.16 5.94 6.63 6.56 5.23
Longitude: 15947 West 21Tilt 5.9 5.82 5.86 5.04 4.56 4.2 4.35 5.07 5.39 5.74 6.11 6.51 5.38
36Tilt 5.19 5.34 5.62 5.08 4.8 4.49 4.6 5.22 5.27 5.35 5.41 6.11 5.21

Issue 1 September 2012 Page 16


Nuku'alofa, Tongatapu, Tonga Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Annual
Average
Latitude: 2108 South 0Tilt 6.69 6.3 5.62 4.65 4.04 3.58 3.78 4.43 5.23 6.28 6.69 6.7 5.32
Longitude: 17512 West 21Tilt 6.1 5.97 5.69 5.1 4.81 4.25 4.41 5.03 5.46 6.08 6.16 6.65 5.48
36Tilt 5.35 5.47 5.46 5.14 5.08 4.55 4.68 5.18 5.34 5.65 5.45 6.25 5.3

Honiara, Solomon Islands Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Annual
Average
Latitude: 0927 South 0Tilt 5.99 5.55 5.61 5.41 4.76 4.59 4.45 5.19 5.81 6.26 6.4 6.22 5.52
Longitude: 15957 East 9Tilt 5.98 5.47 5.54 5.52 5.01 4.91 4.7 5.36 5.82 6.15 6.38 6.24 5.59
24Tilt 5.91 5.29 5.35 5.59 5.27 5.28 4.99 5.53 5.72 5.88 6.28 6.22 5.61

Koror, Palau Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Annual
Average
Latitude: 0720 North 0Tilt 5.19 5.59 6.18 6.3 5.71 5.01 5.12 5.2 5.56 5.39 5.26 4.94 5.45
Longitude: 13428 East 7Tilt 5.4 5.7 6.16 6.22 5.7 5.01 5.11 5.15 5.49 5.45 5.44 5.16 5.5
22Tilt 5.75 5.86 6.06 6.01 5.66 5.03 5.1 5.03 5.3 5.51 5.74 5.54 5.55

Ponape, Pohnpei FSM Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Annual
Average
Latitude: 654 North 0Tilt 4.97 5.57 5.91 5.79 5.44 5.33 5.51 5.54 5.66 5.29 5.03 4.83 5.4
Longitude: 15813 East 6Tilt 5.12 5.65 5.88 5.72 5.42 5.33 5.51 5.49 5.59 5.33 5.16 5 5.43
21Tilt 5.43 5.82 5.8 5.55 5.4 5.39 5.53 5.39 5.41 5.39 5.43 5.36 5.49




Majuro, Marshall Islands



Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Annual
Average
Latitude: 712 North 0Tilt 5.26 5.86 6.11 5.89 5.66 5.31 5.35 5.63 5.42 5.15 4.88 4.84 5.44
Longitude: 17106 East 7Tilt 5.47 5.99 6.09 5.81 5.65 5.32 5.35 5.58 5.35 5.2 5.03 5.05 5.49
22Tilt 5.83 6.16 5.99 5.62 5.61 5.35 5.35 5.46 5.16 5.25 5.27 5.4 5.53


Issue 1 September 2012 Page 17

Alofi, Niue Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Annual
Average
Latitude: 1904 South 0Tilt 6.47 6.2 5.67 4.8 4.26 3.86 4.01 4.61 5.35 6.02 6.53 6.46 5.34
Longitude: 16955 West 19Tilt 6.43 5.88 5.7 5.2 4.96 4.47 4.75 5.14 5.53 5.81 5.98 6.47 5.53
34Tilt 6.06 5.4 5.47 5.24 5.24 4.78 5.08 5.29 5.42 5.41 5.35 6.15 5.41

Nauru Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Annual
Average
Latitude: 032 South 0Tilt 5.77 6.24 6.27 6.04 5.99 5.75 5.85 6.25 6.7 6.5 6.12 5.5 6.07
Longitude: 16656 East 15Tilt 5.94 6.26 6.07 6.06 6.28 6.16 6.21 6.4 6.52 6.45 6.27 5.69 6.19

Vaiaku, Tuvalu Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Annual
Average
Latitude: 831 South

0Tilt 5.16 5.27 5.33 5.29 4.93 4.66 4.76 5.3 5.72 5.8 5.57 5.23 5.25
Longitude: 17913 East 8Tilt 5.14 5.2 5.26 5.37 5.14 4.92 5 5.45 5.71 5.71 5.54 5.22 5.31
23Tilt 5.09 5.05 5.08 5.43 5.42 5.3 5.33 5.62 5.61 5.49 5.47 5.21 5.34

Hagta, Guam Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Annual
Average
Latitude: 1328 North 0Tilt 5.33 5.87 6.73 7.12 7.04 6.44 6 5.3 5.42 5.46 5.16 5.05 5.9
Longitude: 14445 East 13Tilt 5.95 6.27 6.86 6.88 6.97 6.43 5.95 5.06 5.38 5.7 5.66 5.7 6.07
28Tilt 6.41 6.49 6.75 6.4 6.7 6.27 5.76 4.68 5.19 5.78 6.01 6.2 6.05

Noumea, New Caledonia Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Annual
Average
Latitude: 2216 South 0Tilt 7.31 6.7 5.73 4.97 3.94 3.47 3.91 4.73 6.05 7.09 7.41 7.6 5.73
Longitude: 16627 East 22Tilt 6.61 6.34 5.83 5.56 4.76 4.19 4.69 5.51 6.44 6.88 6.77 7.53 5.93
37Tilt 5.75 5.8 5.6 5.63 5.03 4.48 5 5.7 6.33 6.38 5.94 7.03 5.72

Pago Pago, American Samoa Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Annual
Average
Latitude: 1416 South 0Tilt 5.87 5.93 5.54 5.18 4.63 4.4 4.59 5.2 5.78 6.05 6.11 5.93 5.43
Longitude: 17042 West 14Tilt 5.79 5.66 5.51 5.43 5.11 4.99 5.15 5.59 5.88 5.84 6.01 5.87 5.57

Issue 1 September 2012 Page 18

29Tilt 5.57 5.22 5.3 5.49 5.4 5.39 5.52 5.77 5.77 5.46 5.75 5.68 5.53

Monthly Averaged Insolation Incident On A Horizontal Surface (kWh/m/day)
Monthly Averaged Irradiation Incident On An Equator-Pointed Tilted Surface (kWh/m/day)

Source: NASA Surface meteorology and Solar Energy
(http://eosweb.larc.nasa.gov)


Issue 1 September 2012 Page 19



Appendix 1 Table of Abbreviations and Acronyms

d.c. Direct current
a.c. Alternating current
AS/NZS Australia Standard/New Zealand Standard
UL Underwriters Laboratory
ICC International Code Council
NFPA National fire Protection Association
ASCE American Society of Civil Engineers
IEEE Institute of Electrical and Electronics Engineers
Wh Watt hours
kWh Kilowatt hours
W Watts
W
P
Watts peak
H hours
V Volts
A Amps
VA Volt amps
Ah Amp hours
DOD Depth of discharge
C
100
Battery capacity when battery is discharged over 100
hours
PV Photovoltaic
MPPT Maximum power point tracker
PSH Peak sun hours( kWh/m
2
)
kWh/m
2
Kilowatt hours/metres squared
C Degrees Celsius
NOCT Nominal operating cell temperature
T
cell-eff
the average daily effective cell temperature (degrees
Celsius)
T
a.day
the daytime average ambient temperature for the month
that the sizing is being undertaken (degrees Celsius)
f
man
Derating due to manufacturers tolerance
(dimensionless)
f
dirt
Derating due to dirt (dimensionless)
PWM Pulse Width Modulation
V
oc
Open circuit voltage (volts)
V
mp
Maximum power point voltage (volts)
I
sc
Short circuit current (amps)
I
mp
Maximum power point current (amps)
STC Standard test conditions