IRENA

International Renewable Energy Agency

Pacific Lighthouses
Report

Hybrid power systems

Copyright © IRENA 2013
Unless otherwise indicated, material in this publication may be used freely, shared or reprinted, so long as IRENA is
acknowledged as the source.

About IRENA
The International Renewable Energy Agency (IRENA) is an intergovernmental organisation that supports countries in their
transition to a sustainable energy future, and serves as the principal platform for international cooperation, a centre of excel-
lence, and a repository of policy, technology, resource and financial knowledge on renewable energy. IRENA promotes the
widespread adoption and sustainable use of all forms of renewable energy, including bioenergy, geothermal, hydropower,
ocean, solar and wind energy in the pursuit of sustainable development, energy access, energy security and low-carbon eco-
nomic growth and prosperity.
www.irena.org

Acknowledgements
The completion of this reported benefitted from valuable feedback received from ‘Akau’ola (Advisor, Tonga Energy Roadmap),
John van Brink (CEO, Tonga Power Ltd), Thomas Jenson (Energy Adviser, UNDP) and Herb Wade (Consultant).

Authors: Peter Journeay-Kaler (formerly intern at IRENA) and Linus Mofor (IRENA)

For further information or to provide feedback, please contact: Linus Mofor, Innovation and Technology Centre, IRENA.
E-mail: LMofor@irena.org or secretariat@irena.org.

Disclaimer
The designations employed and the presentation of materials herein do not imply the expression of any opinion
whatsoever on the part of the International Renewable Energy Agency concerning the legal status of any country,
territory, city or area, or concerning their authorities or the delimitation of their frontiers or boundaries.
 Pacific Lighthouses

Hybrid power systems

August 2013
 TABLE OF CONTENTS

Preface...................................................................................................................................................................... V

Acronyms................................................................................................................................................................. VI

Summary.................................................................................................................................................................... 1

1. Introduction............................................................................................................................................................2
1.1 E
 nergy challenges and opportunities for renewables in the Pacific islands.............................................................. 2
1.2 E nergy use in the Pacific islands region.............................................................................................................................. 2
1.2.1 Transport sector................................................................................................................................................................................3
1.2.2 Power generation............................................................................................................................................................................3
1.3 Overview of power generation in the Pacific Islands region........................................................................................ 3
2. Diesel power systems............................................................................................................................................6
2.1 M
 eeting demand: active power and frequency control.................................................................................................. 6
2.2 R
 eactive power and voltage control................................................................................................................................... 7
2.3 Spinning reserves...................................................................................................................................................................... 7
2.4 Grid integration of renewable power and its impacts................................................................................................... 8
3. Modelling solar PV integration in the Tongatapu diesel-powered grid............................................................ 10
3.1 Hybrid systems.......................................................................................................................................................................... 10
3.2 C
 urrent power system of Tongatapu...................................................................................................................................11
3.3 Low penetration system..........................................................................................................................................................11
3.4 Medium penetration system.................................................................................................................................................12
3.4.1 Low generator loading................................................................................................................................................................ 12
3.4.2 Variable renewable power output and generator cycling........................................................................................... 14
3.5 High penetration system........................................................................................................................................................15
4. Comprehensive power systems approach.......................................................................................................... 18
4.1 Renewable energy resource data collection.....................................................................................................................18
4.2 Grid modelling and verification...........................................................................................................................................18
4.3 Energy efficiency.....................................................................................................................................................................18
4.4 S
 electing partners with the right expertise.....................................................................................................................18
4.5 Clear steps and measurable goals......................................................................................................................................19
5. Non-technical challenges of high integration of renewables in Pacific Islands diesel-powered grids........... 20
5.1 W
 ide range of affected stakeholders.................................................................................................................................20
5.2 Local capacity building.........................................................................................................................................................20
5.3 L
 egal and regulatory support of renewables..................................................................................................................20
5.4 Island grouping for economies of scale and leveraging of private capital.............................................................21
5.5 Renewables market structure impacts..............................................................................................................................21
6. Conclusion...........................................................................................................................................................22

IV Hybr id p ower s ys te m s
Preface

In the Abu Dhabi Communiqué on accelerating renew- situations, and the challenges and opportunities for re-
able energy uptake for the Pacific Islands (of 13 January newable energy deployment, around the region. These
2012), leaders from the Pacific Island Countries and Ter- studies are available for the Cook Islands, the Federated
ritories (PICTs) called on the International Renewable States of Micronesia, the Republic of Fiji, Kiribati, the
Energy Agency (IRENA) to “…map the Renewable Republic of the Marshall Islands, the Republic of Nauru,
Energy Readiness of the Pacific Islands Countries Niue, the Republic of Palau, Papua New Guinea, Samoa,
and Territories to ascertain the status of renewable the Solomon Islands, the Kingdom of Tonga, Tokelau,
energy opportunities and identify pathways to close Tuvalu and the Republic of Vanuatu. The IRENA Pacific
gaps” and to integrate all IRENA activities in the region Lighthouses report draws on those studies, as well as
“…into a coherent roadmap for the Pacific Islands”. In this additional study on a diesel-renewable energy hy-
response, IRENA has carried out a wide range of activi- brid power system, intended as a transition measure to
ties of specific relevance and application to the PICTs as a renewables-based energy future for the PICTs, which
well as other Small Island Developing States (SIDS). This is also part of the series.
work has now been integrated into the IRENA report:
Pacific Lighthouses: Renewable Energy Roadmapping IRENA, in collaboration with its members and other key
for Islands. development partners, will continue to support the de-
velopment national roadmaps and strategies aimed at
The report consists of an overview roadmap framework enhanced deployment of renewables in the Pacific and
and 15 island-specific studies on the respective energy other island states and territories.

 H y b rid p owe r sy stems V

 Acronyms

GDP Gross domestic product

GWh Gigawatt-hours (Thousand Million kilowatt-hours)

kWh Kilowatt-hours (1 000 Watt-hours)

kV Kilovolts

kVA Kilovolt-ampere

ML Megalitres

MW/ MWh Megawatt (Million Watts)/ megawatt-hours

PV Solar Photovoltaic

RPM Revolutions per minute

USD United States dollar (Currency)

Wp/kWp Watts Peak/kilowatt-peak (Solar PV)

VI Hybr id p ower s ys te m s
Summary

The use of diesel generators to power small and iso- t­echnologies to manage the variability of those
lated electricity grids is especially prevalent in remote sources of renewable energy.
and rural areas in developing economies, especially in (iii) Increasing the level of renewable power integra-
island states and communities where such generators tion also increases the complexity of the systems
are often the sole source of electricity generation. The needed to control variable power output to en-
remoteness of Pacific Islands and the related challeng- sure grid stability, and could require investment
ing and expensive logistics for fuel distribution result in in new diesel generators and improvement of
high fuel costs and low security of supply. These issues grid infrastructure. It is therefore essential that
threaten island energy and economic security and are local technical expertise is in place to ascertain
of particular concern for the many least developed sustainable operation and maintenance of the
countries located in the area. Yet Pacific Islands possess resulting complex grids.
a wide variety of abundant renewable energy resources (iv) Investment in smart meters and load control
that have the potential to greatly reduce or even elimi- systems are needed if islands want to boost RE
nate their dependence on expensive imported fossil penetration through demand side management.
fuels. The associated additional costs must be weighed
against the money saved through reduced fuel
The work reported herein focuses on identifying the consumption. Determining both the system cost
key concepts, challenges and best practices needed to and fuel savings requires a detailed knowledge of
increase the uptake of renewables in Pacific Islands. In the final system configuration and power output.
this context the report shows the results of a case study This requires islands to adopt a comprehensive
assessment and analysis of the potential for solar pho- power systems approach that takes into account
tovoltaic integration into the Tongatapu diesel-based all relevant information and generates a long
power plant in Tonga. The study illustrates the role and term renewable power integration plan with clear
levels that such hybrid systems could play towards a and measurable goals.
transition to a renewables-based energy future for is- (v) Planning efforts for renewable energy de-
lands of the Pacific region, where the power generation ployment need to identify the key skills and
and transport sectors are mostly dependent on expen- knowledge for successful system operation and
sive imported refined oil products. maintenance and determine which skills are not
present in the local labour force. Regional and
The key messages arising from this study are: local policy makers need to engage with Pacific
universities and other institutions to develop lo-
(i) In the medium and long terms, renewables-based cal programs to start building these critical skills.
power solutions would be the most sustainable (vi) With relatively modest investments Pacific Is-
and cost-effective solutions for Pacific Islands lands could soon reach RE penetration levels that
communities. In the transition to that stage, re- will eventually be required worldwide to mitigate
newable power and diesel hybrid systems with climate change. Building strong local capacity in
high levels of renewables integration and energy this area can transform the Pacific region into a
efficiency measures can play a key role in the center of expertise for decarbonising the power
energy supply for island communities and are, sector.
indeed, a viable option for the Pacific Islands. (vii) IRENA is already working with various partners
(ii) Successfully integrating solar power and/or in the regions, including the Pacific Power As-
wind power into a diesel generator-based power sociation, to develop technical expertise for grid
grid requires the use of a variety of enabling stability modelling and management.

 H y b rid p owe r sy stems 1

 1. Introduction

1.1 E nergy challenges and opportunities renewables, mandates that the Agency bears “…in mind
the special needs of the developing countries, and re-
for renewables in the Pacific islands mote and isolated regions and islands.”4
Islands present unique challenges and opportunities for
the deployment of renewable energy (RE). Most Pacific IRENA is demonstrating clear dedication to fulfilling its
islands are located far from major oil distribution hubs islands-specific mandate. One of the first activities of
and depend on complex and lengthy fuel supply chains. the Agency following its official launch in April of 2011
Fuel delivery logistics are often further complicated by was a workshop on accelerated renewable energy de-
lack of modern port facilities, requiring the use of small- velopment, with emphasis on the Pacific Islands region.
er, specialised boats. The small population sizes of many This workshop took place in Sydney, Australia from
islands limit the level of fuel demand while the small ge- the 26th to 28th of October 2011 and brought together
ographic size constrains fuel storage. Both of these fac- leaders and key policy makers from a number of Pacific
tors reduce the purchasing power of such island com- Islands Countries and Territories (PICTs). Discussions fo-
munities. As a result, islands in the Pacific region face cused on identifying key technical, policy, economic and
some of the world’s highest fuel costs and have greater social barriers to the deployment of renewable energy
exposure to price volatility and supply disruptions. In and developing an understanding of how IRENA could
2009 the region’s utilities had consumer electricity tar- assist the PICTs in overcoming those challenges. Grid
iffs that averaged USD 0.44/kWh and for some islands stability was among the key issues identified for further
exceeded USD1.00/kWh. Electricity production costs support. Grid stability is very important if the share of
are likely higher as many Pacific Islands provide subsi- renewables must be increased in existing grids.
dies to protect consumers from the full price of power
generation.1 High energy costs, price volatility and risks This report focuses on identifying the key concepts,
to fuel supply are of particular concern because most challenges and best practices needed to increase RE
Pacific islands have small economies. High cost fuel uptake on Pacific islands through hybrid RE and diesel-
imports can consume significant percentages of GDP, powered systems.
driving up prices on food and other essential items and
limiting investments in education, infrastructure and
other key services. In addition, climate change effects
associated with oil consumption are major concerns for
islands. Pacific islands face a significant threat from ris-
ing ocean levels, with some island having a maximum el-
1.2 Energy use
evation of less than 5 meters above sea level.2 Increased in the Pacific Islands region
storm activity and weather disruptions associated with
climate change threaten the many islands that are in the Transportation and power generation dominate energy
path of seasonal cyclones and typhoons. use in the pacific. Industry is mostly limited to mining
on a few islands and more wide spread agricultural,
Right from its inception, IRENA has acknowledged the forestry and fish processing facilities almost all of which
exceptional status of islands and placed a special focus rely directly on electricity. Traditional biomass use for
on assisting these communities. The opening declara- cooking – although slowly declining in favour of LPG
tion of the Agency’s Statute specifically notes the “… and kerosene – still accounts for the largest share of
the huge potential of renewable energy in providing … overall energy use in rural areas throughout the Pacific.
access to energy for isolated and remote regions and With the exception of some notable contributions from
islands3”. Statute Article IV, which defines the activi- hydropower energy use in the Pacific is dominated by
ties that IRENA will engage in to support the uptake of fossil fuels. The lack of local oil resources and refining
capacity in most of the PICTs means that refined oil
1 Pacific Economic Monitor, Asian Development Bank, July 2010 products must be imported over great distances at high
2 https://www.cia.gov/library/publications/the-world-factbook/ and volatile costs.
geos/kr.html
3 IRENA Statue, Opening Declaration, page 2 4 IRENA Statue, Article IV Activities, page 7

2 Hybr id p ower s ys te m s
1.2.1 Transpor sector 1.2.2 Power generation
Power generation represents approximately 25% of
In the Pacific, transportation accounts for approximate-
energy demand in the Pacific Islands. Hydropower pro-
ly 75% of energy demand and uses almost exclusively
vides a major contribution to electricity generation on
imported refined oil products. A lack of detailed data
Fiji, PNG and Samoa. There is limited use of biomass to
makes it difficult to determine the exact split of fuel
offset power consumption on islands with forestry and
usage between land, sea and air transport. However it
agricultural processing facilities. Solar PV systems, used
can be generally stated that land transport is the largest
widely for rural electrification are spread across the
sector and is dominated by a mix of diesel and gasoline
region. Utility scale wind parks are operational on Fiji
passenger cars and light commercial vehicles. Sea and
and Vanuatu islands. However, the vast bulk of power
air transport can a play notable roles for Pacific states
generation capacity is based on internal combustion
with a wide dispersion of populated islands. Generally
engines and generators utilising imported diesel, heavy
sea transport is the larger of the two with varying con-
fuel oil (HFO) and light fuel oil (LFO).
tributions from interisland passenger and cargo services
and fishing fleets. Local air transit is generally limit to a
small number of light aircraft but can be significant on
islands with developed tourist industries.
1.3 O
 verview of power generation
Currently cost and technical barriers impede a major
shift in the transportation sector from the use of fossil
in the Pacific Islands region
fuels to renewables. However, given that the transpor- Table 1 presents the results of a review of the electricity
tation sector dominates oil consumption in the Pacific generation systems of 15 PICTs. This review was per-
Islands, it is essential that renewable energy-based formed using a variety of sources, including the Pacific
options for transportation be thoroughly examined to Power Association (PPA) 2011 Benchmarking report and
determine when and how they can be deployed on a local utilities. The total generation capacity for the 15
large scale. PICTs is approximately 700 MW. In 2011 approximately

Table 1: 2010/2011 electricity generation statistics

PICT Installed Capacity MW Peak Demand1 MW Annual Generation MW
Cook Islands 10.36 4.9 27,763
FSM-Chuuk 2.0 4.0 9,768
FSM-Kosrae 5.0 1.1 6,504
FSM-Pohnpei 7.6 6.9 38,920
FSM-Yap 6.6 2.3 13,000
Fiji 211.2 139.6 835,169
Kiribati 5.5 5.3 21,641
Marshall Islands-Majuro 28 8.9 75,749
Marshall Islands-Ebeye 3.6 2.0 14,183
Nauru 6.04 3.3 17,103
Niue 3.25 0.54 3,168
Palau 18.9 15.4 84,860
Papua New Guinea 2922 92.94 796,610 + 1,900,0003
Samoa 37.5 18.0 111,353
Solomon Islands 25.6 13.8 83,600
Tokelau4 0.927 0.20 34,000*
Tonga 15.3 7.7 52,609
Tuvalu 5.1 1.0 11,800
Vanuatu (UNELCO) 23.9 11.3 60,360
Vanuatu (VUI)5 4.1 1.71 3,350
Total 712 349 4,201,510
1: Peak demand main island nation grid only, excludes notable power systems on secondary / remote islands & private generation
2: Excludes substantial generation assets dedicated to private mining activities
3: First value: PNG public utility, second value: private mining operations

 H y b rid p owe r sy stems 3

 78% of the power generation on these islands came The small and isolated nature of the majority of Pacific
from generators fueled with diesel, HFO or LFO. The diesel power plants is a particular concern for the in-
remaining 22% was primarily provided by hydropower. tegration of high levels of variable RE. Large intercon-
PNG and Fiji are significant outliers in terms of both total nected mainland grids have a wide variety of generation
capacity and generation mix. Omitting PNG and Fiji, the assets they can draw on to balance out variable RE
total generation capacity is roughly 175MW and consists power generation. Pacific grids are dependent on iso-
almost entirely of diesel generators. Because many lated diesel-powered stations and will have to augment
PICTs are composed of numerous islands, the 175MW their existing generation systems with new technology
is divided among many smaller power plants. To deter- to compensate for variable RE.
mine the characteristics of these individual power plants
IRENA reviewed the 2011 World Electric Power Plants
(WEPP) Database. Table 2 gives an overall picture of Another key concern is the advanced age of the diesel
the diesel generator fleets of the 15 Pacific islands. The generator fleet. The average generator in the Pacific is
database shows that the bulk of the region’s generation around 20 years old. Figure1 shows that the last major de-
capacity consists of power plants of less than 10  MW ployment of generators occurred over ten years ago. This
utilising several generators with capacities ranging from means many generators in the Pacific are approaching or
around 25 kilowatt (kW) to 10 MW. have exceeded the manufacturer’s recommended lifes-
pan. It should be noted that the WEPP database does not
Figures 1 shows a breakdown of the year-by-year and always include direct verification for remote areas such as
cumulative installation of the current diesel generator in the Pacific Islands region. Some of the older generators
fleet. For this figure PNG and Fiji are excluded to give a in Table 2 may no longer be in operation. However, inquir-
better representation of the average conditions on the ies with the PPA identified generators that have been in
smaller Pacific islands. continuous use since the early 1980s.

Table 2: WEPP Pacific island diesel generator statistics (2011)

Unit Size (MW) Operational Year

Capacity Number of
PICTs (MW) Units Min Max Oldest Newest
Cook Islands 11.04 24 0.025 2.1 1990 2009
FS of Micronesia 35.09 36 0.027 3.2 1974 2012
Fiji 67.03 39 0.06 10.15 1953 2011
Kiribati 8.90 9 0.6 1.4 1994 2005
Marshall Islands 41.90 33 0.06 6.4 1982 2003
Nauru 4.00 5 0.8 2002 2005
Niue 1.68 4 0.421 no data
Palau 18.88 17 0.1 3.4 1997 2012
Papua New Guinea 141.34 28 0.14 15 2007 2011
Samoa 16.59 15 0.045 3.5 1979 2001
Solomon Islands 37.78 44 0.04 4.2 1971 2006
Tonga 14.44 19 0.056 1.729 1972 1998
Tokelau no data
Tuvalu 3.71 30 0.045 1 1982 2001
Vanuatu 15.46 14 0.1 4.23 1994 2010
Total 418 317

4 Hybr id p ower s ys te m s
The advanced age of the Pacific diesel fleet has sig- ern diesels. A basic review of diesel generator function
nificant implications when considering high levels of RE and the effects of RE integration are given in Section 2
penetration into existing grids in the region. The older to shed light on the particular challenges posed by the
units may lack computer control systems and likely have small capacity, isolation and advanced age of the Pacific
slower ramp rates and reduced efficiencies versus mod- islands’ diesel fleet.

Figure 1: Year-by-year installation of current diesel fleet (Excludes Fiji & PNG)

 H y b rid p owe r sy stems 5

 2. Diesel power systems

Diesel generators are the most common generation across the island and then stepped back down to a
technology used for power systems in remote areas standard level for distribution to the loads of individual
such as Pacific Islands and often provide isolated com- customers.
munities with their first access to electricity. A number
of factors make diesel generators well suited for island
power generation. Diesel fuel has a high energy density
which greatly reduces fuel storage requirements com-
pared to other fossil fuels such as coal. The generators 2.1 M
 eeting demand: active power
are relatively compact and available in a wide range and frequency control
of capacities allowing for power plants that can be
closely matched to an island’s demand. The majority of A key fact of power generation is that the supply from
power plants in the Pacific use several diesel generators generating units must always match the demand from
running in parallel. This gives operational flexibility as customer loads. Whenever a load is turned on additional
individual generators can be started up or shut down power must be immediately generated. When the load
when demand changes. Multiple small generators also is switched off generators must immediately reduce
increase efficiency and security of supply versus a single their power output. Demand is constantly changing as
large generator. Diesel generators have relatively quick customers switch loads on and off at their convenience.
start times and good flexibility in meeting daily and sea- This can lead to sudden spikes and drops in demand
sonal variations in demand. They provide fast response as major industrial equipment is powered up or shut
times and good power quality and are relatively simple down or when populations arrive at home after work
to operate and maintain. The technology is robust and and power on lights and appliances and then shut these
has a long track record of successful deployment. loads off before going to sleep.

The power that does work on the grid, lighting a bulb

A generator includes a small battery and electrical or powering an oven, is called active power. The diesel
starter motor to bring the main engine online when generators monitor the frequency of the AC electricity
there is a demand for electricity. Once the engine is being produced to sense and respond to changes in
running a small amount of power is used to keep the the demand for active power. Generators are designed
starter battery charged. The fuel system stores the die- to run at certain speed of rotation (rpm) regardless of
sel and delivers it to the engine. The engine burns this their level of power output. A generator’s alternator
fuel, converting chemical energy into thermal energy is designed such that this rpm is directly proportional
(waste heat) and rotational kinetic energy. The engine to the frequency of the AC power being produced. As
is attached to an alternator that converts the rotational a generator is powered on and brought up to speed
kinetic energy into electrical energy. the moving components of the generator, the pistons,
crankshaft and other components acquire a significant
Utility scale generators in the Pacific Islands region amount kinetic energy. These moving parts have a
produce three-phase alternating current (AC) electric- substantial inertia such that if fuel was suddenly cut off
ity, the same type of power generated by traditional from the engine they would continue to spin and their
mainland power plants. Three-phase AC is composed kinetic energy would be transformed into electricity.
of three separate flows of electricity which are equally
spaced apart and oscillate in a sinusoidal fashion. Elec- When a generator is running at steady power out-
tricity grids have three separate sets of wires, one for put and a load on the grid demands more power the
each phase. Electrical loads with high power demand load is able to extract this power instantaneously from
(e.g. industrial motors) often use all three phases while the kinetic energy of the moving components in the
common household loads such as lights and appliances generator. This loss of energy slows the speed of the
use only one phase. The AC frequency is 60Hz in U.S. generator and thereby reduces the frequency of AC
Pacific territories and 50Hz on all other islands. The volt- power output. The controls on the generator sense this
age of the AC electricity produced by generators varies drop in frequency and compensate by increasing the
depending on the model but voltage is usually stepped fuel input to the engine. This raises the power output
up by a transformer to a higher level for transmission of the generator to match the increased demand from

6 Hybr id p ower s ys te m s
the load and returns the generator to its optimal rpm In a simpler alternator the armature uses a permanent
and AC frequency. The opposite is true when a load is magnet with a fixed strength. Advanced alternators in
switched off. In this case, the physical resistance to the diesel generators use an electromagnet in the armature.
generator’s rotation drops when the demand from the The AVR is a set of electronic sensors and controls that
load diminishes. This causes the generator to spin faster is able to use small amount of power from the genera-
resulting in an increasing frequency. The controls sense tor to adjust the strength of the armature magnet’s field
this rise in frequency and reduce the fuel input to the and thereby control the voltage output of the generator.
engine dropping the power output to match the lower
demand on the grid and returning the generator to its When a load requiring reactive power is connected to
optimal rpm and frequency. the grid it causes a drop in voltage as the inductive/
capacitive components inside the load are charged up.
The frequency is not just a control parameter. Maintain- The AVR senses this voltage drop and provides more
ing the frequency within certain limits is also required power to the magnet in the alternator’s armature. This
because the grid and associated loads have all been boosts the strength of the magnetic field and brings
designed to operate within a narrow frequency band. the voltage back to the correct level. When a load con-
Excursions outside this band will result in reduced ef- suming reactive power is removed from the grid the
ficiency and could damage grid infrastructure and at- voltage will go up. The AVR senses this voltage rise and
tached equipment (loads). decrease the power to the armatures magnet, reducing
its field strength and returning voltage the correct level.

Similar to frequency voltage is not just a control varia-

2.2 Reactive power ble. Grid infrastructure and attached loads are designed
to operate in a narrow band of voltages. Voltages on
and voltage control the grid outside of these limits reduce efficiency and
In addition to supplying the active power that performs could cause significant damage. Diesel generator con-
work on the grid, generators must also provide reac- trol systems are designed to monitor and control both
tive power. The amount of reactive power required is frequency and voltage to ensure safe and efficient grid
determined by the number of inductive and capacitive operations. In this manner diesel generators are able to
loads on the grid. Electric motors, transformers and provide power quantity, matching supply to demand,
other devices attached to the grid contain inductive and power quality, maintaining frequency and voltage.
coils and/or capacitors that need to be charged up with
voltage before they can operate. This power to charge
these components is not consumed but stored within
the voltage built up across each component. As such
this reactive power does no work on the grid. However, 2.3 Spinning reserves
it must be supplied by the generator for the grid and
certain loads to function. As loads requiring reactive Another key feature provided by diesel generators is
power come on and off of the grid the generator must spinning reserves. Spinning reserves represent genera-
sense and provide this additional power. tion capacity that is up and running but not providing
power for the grid. This capacity is held in reserve in
While active power control is linked to frequency, reac- case there is a sudden increase in demand at which
tive power control is linked voltage. Although voltage point it can be rapidly deployed to prevent any disrup-
control is commonly supplied by a range of power tion to power supply or quality. Diesel generators are
electronics distributed across the grid, an active volt- designed to operate at a nominal power output but can
age regulator (AVR) on the generator plays a key role in provide significantly higher levels of power output for
maintaining a constant voltage on the grid. Alternators short periods of time. This represents a short term spin-
use the mechanical power from the attached engine ning reserve that can be called on to stabilise the grid
to spin a magnet. This moving section is known as the in case of sudden spikes in demand. In addition diesel
armature. As the armature spins the field of the rotat- power plants commonly keep one or more genera-
ing magnet passes through stationary field coils on the tors running in a standby mode with no power output.
outer wall of the alternator. Each pass of this magnetic Although the fuel to keep this generators running is
field induces an electric current to flow in the field coil, “wasted” because no electricity is being produced it
first in one direction, then the other. This results in oscil- means the full production capacity of the generator is
lating AC electricity that is fed from the field coils onto available as spinning reserves that can be deployed if
the grid. The voltage of this AC electricity is directly demand suddenly increase or in the case of an accident
linked to the strength of the magnetic field passing that damages one of the generators providing power
through the field coils. to the grid.

 H y b rid p owe r sy stems 7

 2.4 G
 rid integration of renewable attention: reduced generator power output (“generator
load”) and the variability of RE power generation.
­power and its impacts
The primary goal of RE integration is decreasing power
Diesel based power plants face a number of challenges output from diesel generators to reduce imported fuel
when viewed in the context of integrating high levels of consumption. High level RE integration can cause gen-
RE. In order to better understand these challenges the erator to run at low load and could even allow the
WEPP data base was reviewed to identify the Original generator to be shut down. Low generator loading and
Equipment Manufacturers (OEMs) of diesel generators frequent stops, however, have important performance
used on 15 PICTs. This data is given in Table 3. OEMs and safety implications.
with substantial presence in the region were contacted
regarding the effects of high RE penetration on genera- Diesel generators operate most efficiently at a certain
tor operations. This process identified two main areas of load, generally 65-80% of the maximum rated capac-

Table 3: WEPP database diesel generator statistics by OEM (2011)

Unit Size (MW) 1st Operational Year

Capacity Units
Engine OEM (MW) No. Min Max Average Old New Average
CAT 40.74 46 0.06 3.25 0.886 1980 2006 1996
Mirlees 28.39 17 0.6 3.5 1.646 1972 2001 1987
Deutz 20.05 10 0.027 6.4 2.005 1995 2000 1999
Crossley 19.27 6 3.2 3.27 3.21 1982 1984 1982
MAN 15.60 12 0.25 4.2 1.300 1962 2001 1975
Cummins 15.49 30 0.103 1.1 0.514 1989 2009 2001
Daihatsu 15.45 8 1.25 2.5 1.931 1991 2004 1998
Wartsila 13.82 8 0.275 4.2 1.728 1988 2003 1995
Mitsubishi 11 3 3.8 4.2 3.6667 1998 1999 1998
Unknown 7.76 55 0.025 1 0.140 1982 2001 1996
Perkins/FG Wilson 6.86 16 0.088 1.5 0.4284 1991 2001 1996
Allen Power 6.60 4 1.2 3 1.65 1987
Duvant 4.20 2 2.1 1991
Enterprise 2.6 1 2.6 1987
Ruston 2.25 3 0.75 1994 1997 1996
Yanmar 2.23 9 0.135 0.35 0.2478 1976 1979 1977
Niigata 1.85 1 1.85     1979
Lister 1.64 9 0.042 0.6 0.183 No Data    
SDMO 1.52 2 0.52 1 0.760 1994 1996 1995
Volvo 1.20 4 0.3 1998
White 0.80 1 0.8 1974
Dorman 0.22 4 0.056 1983
Gardner 0.16 2 0.08 No Data

8 Hybr id p ower s ys te m s
ity. Island power plants are generally designed to meet would need to be quickly restarted. Frequent stop-start
varying demand while keeping generators as close to operation reduces efficiency, increases wear and tear
this load as possible. This delivers higher efficiency and and reduces the useful lifespan of the system. Systems
provides spinning reserves to meet demand increases. will need to be designed to limit start/stop operation.
Operating below this load reduces generator efficiency,
limiting the desired fuel savings. Operation below 30- Low load and start-stop operations are of particular
40% of the rated load can result in engine over-fueling, concern for power systems in the Pacific Islands region
which carbonises injection tips and disrupts the fuel owing to the advanced age of the region’s generator
spray pattern. The resulting poor combustion leads to fleet. Many of these generators have already exceeded
soot formation and un-burnt fuel residue which clogs the OEMs recommend running hours and as such non-
and gums piston rings.5 As a result prolonged low standard operation could severely impact efficiency,
load operation has numerous cost impacts. It further maintenance costs and increases the risk of generator
decreases generator efficiency, increases maintenance failures. Older control systems may simply not allow for
requirements and reduces a generator’s lifespan. prolonged low load operation or rapid shutdown and
startup. There are low load generators that can operate
Low load operation is also a concern for power quality.
safely and efficiently below 10% of rated load. However,
Many RE assets are capable of providing frequency and
the review of PICTs diesel generators indicates that low
voltage control. However, their control systems are nor-
load generators are not currently deployed in significant
mally designed to monitor the frequency and voltage
quantities. In this context it is essential for islands con-
produced by the diesel generator and link their output
sidering high RE penetration to perform a detail inven-
to these “master control” levels. Below 30-40% of rated
tory of their diesel generator assets and factor possible
load most generators cannot support power quality
replacement of current diesel assets into the budgets
on the grid. Technologies exist that will allow some RE
for RE programs.
resources to support power quality on the grid and will
be discussed in detail below. However they are not com-
monly deployed in the Pacific. The variability of RE output is a challenge but it is not
an insurmountable one. Modern generator control sys-
The deployment of technologies such as centralised tems provide a great deal of operational flexibility and
inverters will allow RE generation systems to supply the control systems on older generators can be upgraded.
power quality needed to support grid stability. This will Advanced power electronics and supervisory control
allow generators to be shut down permitting the grid systems can greatly reduce the impacts of RE variability
to run on 100% RE and greatly increasing fuel savings. on diesel grids. The following section examines the tech-
However, if RE production suddenly drops generators nologies and control strategies that allow variable RE
resources to work in harmony with or even completely
5 http://www.generatorsolutions.org/#/gen-faqs/4534128446 supplant diesel power generation.

 H y b rid p owe r sy stems 9

 3. Modelling solar PV integration in
the Tongatapu diesel-powered grid

3.1 Hybrid systems to be shut down instantaneous penetration must exceed

100% – i.e., there must be more than enough RE input
Systems that combine RE and traditional power genera- to cover current demand. Average penetration is a ratio
tion such as diesel generators are called hybrid power of the total RE output (kWh) to the total grid demand
systems. These systems use a wide variety of enabling (kWh) summed over a given time period, normally a
technologies to overcome the difficulties associated month or a year. Average penetration can also exceed
with RE resource variability and allow RE generation to 100% and would represent large surpluses of stored en-
be smoothly integrated into diesel power systems. The ergy for example in a pumped hydro system. Instanta-
possible configurations of hybrid systems are numer- neous penetration relates to hybrid system complexity
ous owing to the wide variety of RE sources and exist- and the required level of control to maintain acceptable
ing enabling technologies. However, hybrid systems power quality. Average penetration gives an indication
can generally be classified – based on the level of RE of the fuel savings and can be used to calculate the
integration they can support – as low, medium or high financial impact of the systems.6
penetration. The U.S. National Renewable Energy Lab
(NREL) has developed definitions for low, medium and
The exact values for RE penetration will vary greatly
high penetration hybrid systems based on its experi-
depending on the type of RE resource and the individ-
ence with wind and diesel hybrid systems. These defini-
ual system configuration but these basic definitions are
tions are detailed in Table 4.
useful for understanding the key challenges of integrat-
ing RE into diesel grids. In order to clearly demonstrate
Here it is important to distinguish between instantane-
ous and average RE penetration. Instantaneous pen- 6 Technology, Performance, and Market Report of Wind-Diesel Ap-
etration is the ratio of the RE power output (kW) to the plications for Remote and Island Communities, Conference Paper,
current grid demand (kW). In order for diesel generators NREL/CP-500-45810 May 2009

Table 4: NREL guidelines for low, medium and high penetration hybrid systems1

Penetration
Penetration Peak Annual
Class Operating Characteristics Instantaeous Average
Low ●● Diesel(s) run full-time
●● Wind power reduces net load on diesel
<50% <20%
●● All wind energy goes to primary load
●● No supervisory control system
Medium ●● Diesel(s) run full-time
●● At high wind power levels, secondary loads dispatched to ensure 20%-
50%-100%
sufficient diesel loading or wind generation is curtailed 50%
●● Requires relatively simple control system
High ●● Diesel(s) may be shut down during high wind availability
100%- 50%-
●● Auxiliary components required to regulate voltage and frequency
400% 150%
●● Requires sophisticated control system

1: http://www.akenergyauthority.org/wind/02_Wind-diesel_power_systems_basics_01-01-2008.pdf

10 Hybr id p ower s ys te m s
hybrid system function and how these systems can 3.3 Low penetration system
assist Pacific Islands in achieving high levels of RE in-
tegration, a step by step example is given below taking Tongatapu Installed a 1MW PV system in July 2012
Tonga’s main island of Tongatapu from low to medium (Maama Mai solar farm). This level of RE integration rep-
to high levels of RE penetration. resents a low penetration hybrid system as it requires
no special control equipment. Figure 3 shows that the
PV generation will have a limited impact on the diesel
generators, although the Maama Mai solar farm will pro-
vide for only around 4% of Tongatapu’s total electricity
3.2 C
 urrent power system demand. The PV output is seen as a negative load by
the generators, which continue to match their output to
of Tongatapu the changing demand profile and support power quality
The Tongatapu’s power plant in 2012 consisted of seven on the grid. The gap between the reduced diesel output
diesel generators with a total maximum rated capacity and grid demand represents the fuel saved over one
of 11.28MW. Detailed statistics on the power plant are day of operation. Inverters on the PV system monitor
given in Table 5. Figure 2 shows a typical weekday de- the frequency and voltage on the grid and match the
mand curve for Tongatapu and demonstrates how the system’s frequency and voltage to these values. Even
diesel generators are capable of easily matching their a 100% drop of the PV system output at the noon peak
output to the grid’s demand for power. represents less than 10% of generator capacity and

Table 5: Tongatapu power plant statistics (2011)

Gen- Fre- Base
erator Engine Engine Se- Alternator Alternator quency Voltage Base Max Power Speed Year in
Name Model rial No. Model Serial No. (Hz) (kV) kVA (kW) (rpm) Service

SR4HV
CAT
UNIT 1 7RN00540 Kato Engi- 8JS00812 50 11 1,525 1,400 1,500 1998
3516B
neering

SR4HV
CAT
UNIT 2 7RN00541 Kato Engi- 8JS00814 50 11 1,525 1,400 1,500 1998
3516B
neering

SR4HV
CAT
UNIT 3 7RN00542 Kato Engi- 8JS00809 50 11 1,525 1,400 1,500 1998
3516B
neering

SR4HV
CAT
UNIT 4 7RN00543 Kato Engi- 8JS00809 50 11 1,525 1,400 1,500 1998
3516B
neering

SR4HV
CAT
UNIT 5 7RN00544 Kato Engi- 8JS00813 50 11 1,525 1,400 1,500 1990
3516B
neering

SR4HV
CAT
UNIT 6 7RN00989 Kato Engi- 8JS00838 50 11 1,525 1,400 1,500 1998
3516B
neering

LSA56
CAT MaK B2 11-10P
UNIT 7 38232 600374-1 50 11 3,456 2,880 600 2006
6CM32C Leroy
Somer

CAT MaK No No
UNIT 8 No data No data No data No data No data No data Q1 2012
6CM32C data data

Trailer CAT Out of

No data No data No data No data 11 1,525 1,400 1,500
Unit PM3516B service

 H y b rid p owe r sy stems 11

 8

Demand
7
Diesel
6

Load (MW)
4

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Time (hours)

Figure 2: Tongatapu typical weekday demand and diesel power

output

could be covered by diesel spinning reserves. This ex- There are numerous strategies to prevent low generator
ample helps to show that Pacific islands can commence loading, thus allowing higher levels of variable RE onto
with smaller RE projects without major effects on their the grid. One common and easy to implement approach
current generation assets. is power limiting. In this situation control equipment
monitors generator load and when high PV output risks
reducing generator load below the lower loading limit
the control equipment has the inverter on the PV system
reduce the power output. This prevents the generator
3.4 Medium penetration system from dropping below the lower limit protecting it from
poor combustion and loss of power quality. Figure 5
In order to achieve significant reduction in diesel fuel shows the same 5MW system and demand profile with
consumption it will be necessary for Tongatapu to install power limiting systems installed.
additional PV capacity. Figure 4 shows the notable drop
in diesel generator load that could be achieved with the
deployment of a 5MW PV system. This would be classi- While power limiting protects diesel generators from
fied as a medium penetration system because it will re- low loading it wastes the additional PV power available
quire that additional control systems to work effectively at the noon peak and reducing the fuel saving delivered
with the existing diesel generators. There are two key by the PV system. Even with power limiting, the 5MW
reasons why additional control equipment are required system still delivers significantly more fuel savings than
– protecting the diesel generators from low loading and the 1MW system. However the additional fuel offsets
limiting generator cycling. require additional control systems that increase the cost
of the PV system.

Additional options exist to deal with low loading. Instal-

3.4.1 Low generator loading
lation of dump loads can turn RE overproduction into
As previously noted diesel generators have a lower load useful energy (e.g. heat). The limited heating demand
limit of approximately 30-40% of rated capacity. Below in the Pacific constrains the value of this option. There-
this limit the generator suffers from poor combustion fore load dumping for ice making for cooling demand
that reduces efficiency, increases maintenance cost and is an option for dealing with low loading. Specialised
can cause permanent damage that reduces the gen- low load diesel generators that are designed to operate
erator’s usable life span. In addition, below this lower efficiently and provide power quality down to 10% of
loading limit the generator can no longer support power rated capacity are an attractive option for Pacific islands
quality on the grid and there is a risk of damaging grid with older generators but would represent a substantial
infrastructure and attached loads, and even of causing cost for islands where the generators still have usable
a black out. lifespan. The cost of each of these options will have to

12 Hybr id p ower s ys te m s
 8

Demand
7
Diesel
6
PV
5
Load (MW)
4

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Time (hours)

Figure 3: Tongatapu demand and diesel power output with

1MW PV

Demand
7
Diesel
6
PV
5
Load (MW)

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Time (hours)

Figure 4: Tongatapu demand and diesel power output with

5MW PV

Demand
7
Diesel
6
PV
5
Load (MW)

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Time (hours)

Figure 5: Power limiting of 5MW PV system to prevent low

generator loading

 H y b rid p owe r sy stems 13

 be weighed on a case-by-case basis against the value of which time the stored energy is released into the grid.
offset fuel consumption. In this way the output from the PV can be maintained
at a relatively constant rate greatly reducing generator
cycling and the associated negative effects.
3.4.2 Variable renewable power output
and generator cycling A variety of commercially available systems are capa-
ble of providing short-term energy storage. The most
An additional problem for medium penetration systems
common are mechanical flywheels (which store RE
arises from the variable RE power output. Figure 6
overcapacity as kinetic energy in spinning mass that can
shows 15 minutes of PV output near the noontime peak
be easily convert back to electricity) and certain batter-
and demonstrates how typical RE output variations can
ies such as Li-ion that have rapid charge and discharge
cause rapid generator cycling even with a flat demand
rates. Regardless of the technology chosen the energy
profile.
storage and associated control equipment represent an
additional cost. In addition, while these technologies are
Generator cycling reduces efficiency and can cause
capable of stabilizing short-term low magnitude varia-
damage that increases maintenance costs and reduces
tions in RE output major drops in RE power will have to
the usable lifespan of the generator. More significantly
be covered by diesel spinning reserve. Maintaining the
the RE variability can exceed the response ramp rate of
high level of spinning reserves needed to back up large
the diesel engine leading to mismatches in power sup-
drops in RE output reduces the fuel savings for medium
ply and demand that can destabilise the grid and cause
penetration hybrid systems. All these cost must be bal-
a black out. As with low generator loading there are
anced against the potential fuel savings to determine if
additional technologies that can be added to the power
the additional investment is worthwhile.
generation system to limit the effects of RE variability.

The primary technique utilised to limit generator cycling The additional control systems required for medium
and protect grid stability in medium penetration hybrid penetration hybrid systems are largely automated and
systems is short-term energy storage. Figure 7 exam- the operation and maintenance (O&M) of these systems
ines the same 15 minutes of noontime PV peak output can likely be supported by the current skill set of Pacific
to illustrate how short-term energy storage is used to utilities. Medium penetration systems have been suc-
reduce the impact of variable RE output and could re- cessfully demonstrated on numerous systems through-
duce diesel costs by up to 50%. With short-term energy out the world mostly to support wind energy systems
storage a control system uses a variety of factors to set although several PV systems have been deployed in
a nominal output for the PV system and monitors the Australia.7
real time output to look for significant fluctuations that
could exceed the response ramp rate of the generator. The main consideration for medium penetration hybrid
PV output above the set point is diverted from the grid power systems is the tradeoff between the costs of the
into the energy storage system. This energy is stored
until the control system detects drop in PV output at 7 http://www.apva.org.au/

Demand
7
Diesel
6
PV
5
Load (MW)

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Time (minutes)

Figure 6: PV variability and diesel generator cycling

14 Hybr id p ower s ys te m s
 Figure 7: Short-term energy storage function & reduced gen-
erator cycling

additional systems needed to boost RE penetration generator to match supply to demand and provide the
and value of the associated fuel savings. The high cost frequency and voltage needed to stabilise the grid. In
of diesel fuel in the Pacific Islands means that medium the scenario shown in Figure 8 the grid would become
penetration hybrid systems are likely to be cost effec- unstable as the generators dropped below their lower
tive. While system O&M is relatively simple, the system load limit and a black out would occur if the generators
design process is quite complex and needs to be pre- were shut off. Power limiting could be used to reduce PV
cisely executed to ensure that the full potential for fuel input and keep the generators above their lower limit.
reductions are achieved. High quality site-specific RE While this would help to stabilise the grid, the wasted
resource data is essential for this design process and the PV overcapacity would likely make it very difficult to
lack of such data is a major impediment to widespread achieve a positive return on the financial investment
deployment of medium penetration systems in the Pa- in the larger PV system. In order for high penetration
cific Islands region. systems to function additional technology must be
installed on the grid that allow the available RE power
output to be matched to demand and support power
quality while the diesel generators are shut down.

3.5 High penetration system A number of technologies including synchronous con-

While a successfully operating 5MW PV medium pene- densers, load banks, dispatchable loads, power convert-
tration hybrid system would considerably reduce diesel ers and advanced system controls can be employed
fuel consumption on Tongatapu it would most likely not to support demand matching and power quality while
achieve the 50% annual reduction outlined in Tonga’s the diesel generators are shut down. Additionally de-
energy roadmap. This level of fuel saving will proba- mand-side load management, including load-shedding
bly require a high penetration hybrid system that can schemes, can be utilised to support grid stability and
provide sustained 100% instantaneous RE penetration reliability. However one of the most promising technolo-
allowing the diesel generators to be completely shut gies to enable high penetration hybrids systems is long-
down for long periods during the day. To illustrate how term energy storage.
such a system could function, Figure 8 shows the effect
of an 8MW PV system on Tongatapu’s diesel generators. Long-term energy storage can be provided by a number
of technologies. For most Pacific Islands large battery
Figure 8 shows that the 8MW PV system can achieve banks are the most likely form of long-term energy
over 100% instantaneous penetration for roughly 4 storage. While lead-acid batteries have been used suc-
hours around the noontime peak. This could allow the cessfully in small capacity high penetration hybrid sys-
diesel generators to be shut off. However, even with tems their relatively low energy density and concern
control systems for a medium penetration hybrid sys- over toxic materials limit their potential deployment on
tem in place the grid is still dependent on the diesel islands. A number of advanced battery chemistries offer

 H y b rid p owe r sy stems 15

 promising performance but NaS batteries are the only PV-based high penetration hybrid systems have been
commercially available battery with the high energy successfully deployed on smaller Pacific islands such
density, rapid discharge rate and non-toxic composi- as Apolima in Samoa, where a 13.5 kW PV system with
tion suitable for high penetration hybrid island power batteries has successfully provide 24-hour power since
systems.8 Figures 9 and 10 give an illustration of how 2006, eliminating the use of diesel power generation
a battery based high penetration hybrid system could and eliminating the importation of 2,460 liters of diesel
allow for 8MW of PV to be successfully operated on per year. On the Japanese island of Hokkaido, a grid
Tongatapu. connected demonstration project with 5MW PV and
1.5MW of NaS batteries has been successfully operated
In Figure 9, when PV output reaches 100% instantane- for over four years.9
ous penetration the diesel engine is able to shut down.
PV production exceeding the current demand is di- The example given in Figures 9 and 10 greatly simpli-
rected into a large capacity battery. Additional control fies the operation of high penetration hybrid systems,
equipment such as a multi-MW centralised inverter which require complex control systems to simultane-
takes over the role of matching power supply (now ously monitor and operate a wide variety of technolo-
coming exclusively from PV) to current demand and gies. Even in this case the diesel generator is still running
provides the appropriate frequency and voltage on the for the majority of the day. Increasing the capacity of
grid. If a major drop in PV output occurs the battery the PV system and battery to provide 24-hour power is
will act as a buffer providing the needed energy to the financially prohibitive for larger islands like Tongatapu.
centralised inverter. If the drop in PV is sustained over a As such further fuel savings will need to come from in-
long period the battery and inverter will give the control tegrating other RE resources into this already complex
system time to bring the diesel generator back online system. Wind is an excellent complement to solar as it
to support the grid. The sustained midday PV over generally blows most strongly at night when PV output
production allows the battery to be charged to a high is zero. However, wind output is generally much more
level. This energy can be used later in the day as shown dynamic that PV, thus increasing the level of control
in Figure 10. required to stabilise high levels of integration. Detailed
resource and power system modeling are required to
confirm the profitable operation of the complex sys-
Figure 10 shows how the energy stored in the battery tems.
can be deployed in several ways. By the end of the
midday PV peak the battery has attained a high state Successful deployment of high penetration hybrid sys-
of charge. As PV production drops off this energy can tems is required to achieve many of the RE targets set
be deployed to keep the diesel engines shut down for a by Pacific Islands but significant work needs to done on
longer period of time. The remaining battery charge can RE resource data collection and power system modeling
be slowly depleted to keep the diesel running at a lower before such systems can be widely deployed. In addi-
level. As the PV outputs drops to zero the centralised tion, proper O&M of the complex control systems and
inverter can hand control over matching demand and numerous enabling technologies will require significant
power quality back to the diesel generator and save the training of the local workforce. All of these challenges
remaining battery charge to offset most of the evening can be overcome but they require Pacific Islands to take
peak. In this manner large additional fuel reductions can a comprehensive approach to power systems planning.
be achieved.
9 http://www.fasid.or.jp/daigakuin/sien/kaisetsu/gaiyo21/pdf/15-4.
8 http://www.credp.org/Data/YounicosRES_EN.pdf pdf

16 Hybr id p ower s ys te m s
 8

Demand
7
Diesel
6
PV
5

Load (MW)
4

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Time (hours)

Figure 8: Tongatapu demand and diesel power output with

8MW PV

Demand
7
Diesel
6
PV
5
Battery
Load (MW)

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Time (hours)

Figure 9: Long-term energy storage matching RE output to

midday demand

Demand
7
Diesel
6
PV
5
Battery
Load (MW)

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Time (hours)

Figure 10: Long-term energy storage shifting PV overcapacity

to supplying evening peak

 H y b rid p owe r sy stems 17

 4. Comprehensive
power systems approach

4.1 Renewable energy highly dynamic nature of small island grids the equip-
ment required for these measurement needs a have
resource data collection a very high sampling rate and represents a non-trivial
The first step in a comprehensive power systems ap- expense.
proach is a detailed review of locally available RE re-
sources. The underlying RE resources are the key driver
for the configuration and power output of the system
and need to be understood in detail. Undoubtedly, a
number of RE studies in the Pacific Islands region have 4.3 Energy efficiency
been carried out by various development and other or-
Static modeling will help to identify grid related energy
ganisations. The data from these reports would provide
efficiency opportunities but all island RE planning ef-
a great resource if collated. As such island RE planning
forts need to go beyond this and include a comprehen-
should start as soon as possible with an effort to identify
sive examination of opportunities for increasing energy
key RE production sites and begin data collection to
efficiency across the island. Energy efficiency efforts
verify resource availability.
such as replacing inefficient electric heaters with solar
thermal water heaters or eliminating the use of incan-
descent light bulbs can lead to notable reductions in de-
mand thereby allowing a smaller and less expensive RE
4.2 Grid modelling and verification system to cover a higher percentage of island demand.
Many energy efficiency efforts can be rapidly deployed
The need to integrate RE systems into existing grids and leading to immediately reduced diesel consumption
the notable impact this can have on diesel generators that frees up additional funding for RE projects.
and grid stability requires that islands have a thorough
understanding of current grid function. This entails a
thorough examination of local demand characteristics
and generation assets along with an assessment of
the physical and electrical grid structure. Once this 4.4 Selecting partners
information has been collected, detailed static and dy-
namic modeling should be performed. Static modeling with the right expertise
examines power system performance under a constant
load and is used to locate grid losses and inefficiencies. Detailed knowledge of RE resource availability and the
If detected these issues will need to be corrected so results from grid modeling and verification will help to
that the full value of RE production can be delivered to define the baseline options for an island’s hybrid power
customers. Dynamic modeling allows system designers system. However even with these details established
to examine how possible disruptions from RE genera- the wide variety of RE technologies and diversity of
tion affects grid stability. In addition to detailed knowl- associated power electronics, storage technologies and
edge of the grid and attached loads, effective dynamic other equipment required to control RE power pro-
modeling requires knowledge of the RE resource. This duction results in a large number of possible system
allows identification of the key effects of potential RE configurations.
technologies on grid stability and drives the selection
of the additional technology best suited to allow RE Sorting out which system is most efficient, economi-
integration. cal and best fits an island’s power needs requires the
expertise of a power system designer with established
A number of software packages exist that can perform experience in field of hybrid systems. In addition, is-
static and dynamic modeling. However, more important land-specific challenges such as limited infrastructure,
than which package is chosen is the selection of a mod- high temperature, humidity, salinity, strong storms and
eler who has intimate familiarity with the operation of other key factors need to be accounted for in the design
smaller diesel based grids. All model results need to be and selection of components. This makes the selection
directly verified with onsite measurement. Due to the of designers and system installers with island experi-

18 Hybr id p ower s ys te m s
ence and knowledge of local conditions an important identify particular challenges and prepare for the up-
factor in successful system deployment. coming phases of project implementation. This com-
prehensive approach can identify which project options
Island RE project planners should place a high priority have a positive return on investment and guards against
on identifying experienced project partners. Given the costly mistakes that are a particular concern given high
limited Pacific RE systems deployment, island RE proj- RE system cost and limited access to project funding in
ect planners need to make sure off-island experts are the Pacific Islands region.
accounting for island specific challenges.
A comprehensive power systems approach is essential
to widespread RE deployment in the Pacific. However,
IRENA’s review of RE projects to date and consultations
with local utilities and policymakers have identified a
4.5 Clear steps and measurable goals number of critical policy and financial challenges that
The technical complexity and numerous processes re- will have to be addressed before RE power systems
quired to deploy RE systems that can reduce island oil can be widely deployed in the Pacific. It is essential that
dependence can be overwhelming. RE planning efforts the project planning process be expanded beyond the
should take a long-term view and develop a detailed technical design realm and include local experts with
understanding of the specific steps required to achieve knowledge of the Pacific’s unique policy and financial
high levels of RE penetration. This step-by-step process environment. To assist in this process some of the key
needs to include clearly defined and measurable goals non-technical challenges identified by IRENA are de-
so that RE project planners can ascertain their progress, tailed below.

 H y b rid p owe r sy stems 19

 5. Non-technical challenges of
high integration of renewables in
Pacific Islands diesel-powered grids

5.1 W
 ide range of fuels require the efficient operation of a wide range of
advanced technologies which have different operation
affected stakeholders constraints and maintenance needs. RE planning efforts
High penetration RE projects have impacts on Pacif- need to identify the key skills and knowledge for suc-
ic communities that extend beyond those associated cessful system O&M and determine which of these skills
with the small footprint of diesel based power systems. are not present in the local labour force. Regional and
The high land use of many RE technologies is clearly a local policy makers need to engage with Pacific univer-
challenge for small islands where space is at a premium. sities and other institutions to develop local programs to
Intricate land ownership regulations and numerous cul- start building these critical skills.
tural sites can greatly complicate and delay landuse-in-
tensive RE and power transmission projects. The visual A local high skills labour force will not only increase the
impact of some RE technologies can create significant capacity for O&M of high RE systems; it will also allow
community resistance, especially in areas dependent on the Pacific communities to expand the benefits of these
“pristine” tropical island environments to support tourist beyond energy security. Detailed knowledge of how to
industries that often represent significant percentages successfully operate high penetration RE systems is of
of national GDP. These concerns and many others mean incredible value across the globe. With relatively modest
that a wide variety of people will be directly affected by investments Pacific islands can soon reach RE penetra-
large RE systems. As such, social acceptance of RE is tion levels that will eventually be required worldwide to
key and it is essential that the planning process include mitigate climate change. Building strong local capacity
a broad base of stakeholders and well-executed public in this area can transform the Pacific into a center of
information campaigns. Identifying and addressing key expertise for decarbonising the power sector.
community concerns upfront prevents costly changes
to projects and the creation of unnecessary opponents.
Broad stakeholder participation is especially important
given that off-island technical experts without direct
knowledge of the communities concerns are likely to be 5.3 L egal and regulatory
the designers of many projects.
support of renewables
On most Pacific Islands power is provided by a single
utility, typically state owned. However, a high incorpora-
5.2 Local capacity building tion of RE will likely requires numerous privately owned
RE systems including distributed RE on private homes
IRENA’s review of Pacific RE projects to date, primarily and larger systems built by investors and project devel-
kW scale PV, revealed that the lack of proper mainte- opers for return on investment. The legal and regulatory
nance was one of the major causes of underperfor- structures on most islands have been designed to sup-
mance and system failure. Pacific islands now have a port a single utility and as result can pose notable chal-
much greater familiarity with these systems and have lenges to incorporating RE. In some cases the state util-
learned that well-trained technicians from local utilities ity is the only entity that is legally allowed to generate
should handle O&M. The current level of power systems power. More common is a lack of the legal frameworks
expertise in the Pacific can most likely support a sig- needed to support the entry of independent power
nificant buildup of low to medium penetration PV and producers (IPP). For example it may be unclear if the
wind systems. However high penetration RE systems law allows power-purchasing agreements with a private
that greatly reduce or eliminate the importation of fossil utility or defines who in government is responsible.

20 Hybr id p ower s ys te m s
Another common issue is electricity subsidies. The sub- the Pacific Islands region. The considerable capital cost
sidy structure on many islands is designed to protect of RE systems will place a heavy burden on these limited
customers from volatility and high prices of diesel. Sig- resources. As such it is essential that all efforts be made
nificant uptake of RE will likely require a legal rework to to reduce RE project costs. A promising possibility is the
allow some of the money to be dedicated to RE. Islands creation of regional islands group based on common
will have to decide what type of subsidy scheme to characteristics such as the type of RE resources avail-
adopt, for example feed in tariffs versus RE production able. These groups could collaborate to combine their
certificates and will have to write and enforce new regu- RE project planning to create a large demand for design
lations to supports these subsides. In addition the cur- services, equipment installation and workforce training.
rent codes and standards will likely need to be updated This would create economies of scale and increase the
to allow RE generation technologies, associate power chances of leveraging private capital.
electronic and energy storage systems to be connected
to the grid.

5.5 Renewables market

structure impacts
5.4 Island grouping for economies
High penetration RE systems will likely result in a com-
of scale and leveraging of plex market structure; with power provided to the grid
private capital by public and private utilities, and consumers with small
RE generation systems. The regulations controlling this
Public funding from development banks, NGOs, multilat- market will need to ensure both affordable electricity
eral institutions and mainland governments of territories prices and sufficient return on investment of RE pro-
currently dominate the financing of RE deployment in jects.

 H y b rid p owe r sy stems 21

 6. Conclusion

This review of RE deployment potential in the Pacific skills. Building this local capacity also represents an op-
uncovered several key messages. Despite the wide di- portunity for the Pacific to become a global center of
versity among Pacific islands they share a common expertise in hybrid power systems. IRENA is assisting
dependence on high cost imported oil. Consequently the Pacific islands in local capacity building by working
power generation from a variety RE resources are with the PPA to boost its ability to perform dynamic
price competitive with diesel generators that currently power systems modeling, a key skill required to deploy
dominate the region’s power sector. A review of Pacific RE power generation.
RE resources has determined that a wide variety of RE
resources are available across the Pacific with the long-
A significant buildup of RE generation in the Pacific is
term potential to greatly reduce or even eliminate the
inhibited by the limited financing in the region, which
region’s oil dependence. However, in the near-term
is primarily composed of public funding from various
most of these RE resources are limited to niche roles.
development partners. The creation of island groups
As such PV and wind are the RE technologies that can
with similar RE resource presents a potential opportu-
be most widely deployed in the near future. However,
nity to create the economies of scale required to bring
energy efficiency measures can play a key role in the
down costs and provide the needed access to private
energy supply for island communities and are, indeed, a
financing. Working with NREL, IRENA is providing Pa-
viable quick-win option for the PICTs
cific islands with information on island groupings and
other options to increase access to private RE project
The variable nature of PV and wind power generation
financing. Furthermore, the United Arab Emirates has
and the dependence on diesel generators mean that
offered a commitment from the Abu Dhabi Fund for De-
high levels of RE utilisation will require complex hybrid
velopment (ADFD) of up to USD 350 million to support
powers systems. As such, island RE planners need to
financing of renewable energy projects in developing
adopt a comprehensive energy systems planning ap-
countries endorsed by IRENA. The IRENA/ADFD pro-
proach that examines all the entire process for deploy-
ject facility is currently evaluating projects from various
ing high penetration RE systems and delivers a clear
countries, including countries in the Pacific Islands re-
step-by-step process with well-defined measurable
gion, for the first round of funding from the fund.
goals. The first step in the process, direct verification of
site specific RE resource availability needs to be rapidly
undertaken due to long data collection periods and the Finally it needs to be noted that although the focus of
strong impact of RE resources on system performance this report is on near term deployment of RE power
and financing. IRENA is actively working to assist Pacific generation this sector represents on average only 25%
Islands in this task with the development of a global RE of Pacific island oil imports. Achieving reduced depend-
resource atlas. ence on fuel imports will require regional RE planners to
address the more challenging issue of wide spread RE
The involvement of policy makers to support local ca- use in the transport sector. Electric vehicles are the best
pacity building is essential for project success, as high technology currently available for RE based transport
penetration RE systems will require training a local but require deployment of large RE power generation
workforce with a wide variety of specialised technical to offset fuel consumption.

22 Hybr id p ower s ys te m s
IRENA
C67 Office Building, Khalidiyah (32nd) Street
P.O. Box 236, Abu Dhabi,
United Arab Emirates
www.irena.org

Copyright © IRENA 2013