Identifying risks, costs, and lessons from ARENA‐funded off‐grid renewable energy projects in regional Australia

The Australian Renewable Energy Agency (ARENA) has provided grant funding to 18 off‐grid and fringe‐of‐grid renewable energy projects under the Regional Australia's Renewables (RAR) program since 2013. This program was designed to address real and perceived risks associated with early stage, precommercial renewable energy development and provide a foundation of demonstration projects to enable the development of a competitive renewable energy sector. These projects range from low to high renewable energy fractions at megawatt scale in remote regions of Australia and encompass a variety of sectors, such as mining, tourism, and remote communities. All projects use photovoltaic as a key technology, often supplemented by additional technologies. The experience from these projects shows that land acquisition, technical integration, stakeholder engagement, and access to finance are among the main reasons for project delivery delays. A qualitative assessment for the remoteness premium is given, based on a comparison of ARENA‐funded on‐grid and off‐grid projects. This indicates that the structural barriers of governance, supply chains, and finance need to be tackled further to lower soft costs. One of the key enablers for future lower renewable energy costs is ARENA's Knowledge Sharing model, through which the funding agency is recompensed by data and information that is provided to the market and increases the impact of ARENA funding.


| INTRODUCTION
Globally, estimates for the number of people with limited or no access to reliable (on-grid) electricity are more than 1.2 billion. 1,2 Predominantly, these people live in high-insolation areas, such as sub-Saharan Africa and India. 2 In this context, the potential exists for a rapid cost-effective electrification with low(er) carbon dioxide emissions using photovoltaic (PV) power as the key low-emission generating technology.
The sustained cost declines observed worldwide for grid-connected PV and reported on using learning curves, [3][4][5][6] together with high fossil fuel-powered generation costs in remote and regional Australia, have resulted in revived interest in the use of renewable technologies for electric power supply in these areas: Approximately 2% of Australia's population lives in remote and regional Australia as classified by the Australian Statistical Geography Standard, 7 yet are in locations with some of the highest renewable energy resource potential-in particular solar power.
The people and businesses in remote and regional Australia account for circa 6% of the country's annual electricity use, with a 5:1 ratio of energy use of mining to remote communities. 8 Diesel and natural gas-powered generators dominate the Australian off-grid electric power supply, with limited renewable power capacity, particularly within larger mines and communities. The gap between what is then identified on paper as potential for PV applications and what is installed in practice remains large. It was within this context of rapidly improving economics of solar and wind, and the expectation that these technologies were now cost competitive in many parts of regional Australia, that the Australian Renewable Energy Agency (ARENA) launched The RAR program was developed to stimulate renewable energy uptake in remote areas with 2 streams: one focused on industry such as mines and tourism facilities and another focused on remote communities and islands. The industry-focused stream included funding for fringe-of-grid projects, which are connected to a centralised grid, but are located in weak areas of the network, such as at the end of very long transmission lines, where diesel backup is often required to maintain power supply. Grant funding under the RAR program was designed to cover a part of the total project cost (the remainder being funded by equity and/or debt by the proponent) and was aimed at bridging the (perceived or apparent) risk gap and/or improve the financial viability of the project. A critical difference of the RAR program to similar funding in other countries is ARENA's Knowledge Sharing requirement, which obliges project proponents to provide ARENA with information. This includes • Standard project information (system size, technologies used, application purpose, expected renewable energy penetration and/or fossil fuel displacement, …).
• Financial models, including the impact of the ARENA grant on project finances.
• Operational data post completion for a period of up to 3 years.
• Lessons learnt (Knowledge Sharing) reports for external distribution to the market.
Through this, ARENA receives a wealth of objective and verified information on projects that is otherwise difficult to obtain, which can help to answer key questions relating to the uptake of renewables in the off-grid and fringe-of-grid space: "What is the cost premium for remote locations (the 'remoteness premium') when compared to on-grid projects?", "What is the risk appetite by key stakeholders (users and financiers) for such projects?", and "How do first-mover costs and integration challenges evolve over time?". Furthermore, the data provided allows ARENA to benchmark future investments, to ensure that its funding continues to stimulate innovation throughout the value chain. The intended goal of ARENA's Knowledge Sharing on RAR-funded projects is to accelerate the impact of its funding and increase awareness in stakeholders, so that future costs can be reduced over time and that risks can be identified, quantified, and, where possible, managed or reduced.
The terms renewable energy fraction (REF period ) and renewable power fraction (RPF) are used in this article instead of renewable penetration, as they allow better clarity for communication regarding how renewable power and energy serve the load and how both the generation and the load must be considered when designing, or discussing, power systems where a renewable generator is included. Ambiguity around the term renewable penetration is removed by describing characteristics using both the REF and the RPF metrics.
In both Equations 1 and 2, the subscript Ren is for renewable energy or power, respectively, where the load is the consumption of electricity by the This article discusses the RAR program as a potential template for other funding bodies such as governments, multilateral development banks, or nongovernmental organisations (NGOs). Some of the key risks encountered throughout the RAR program and its funded projects will be discussed, together with lessons learnt by ARENA, the funding recipients, and stakeholders. The costs and project delivery models of the RAR-funded projects are used to illustrate the structural barriers 9 encountered in supplying renewable power in remote locations, such as remote and regional Australia. Lastly, the RAR program serves as a "sandbox" in which components of the structural barriers can be tackled or elucidated, to achieve high renewable energy fractions on the grid (isolated or not), of which certain aspects will increasingly manifest themselves in the coming years. 10 2 | REGIONAL AUSTRALIA'S RENEWABLES PROGRAM OVERVIEW

| Historical context in Australia
ARENA became active on 1 July 2012 with the mandate to increase the uptake of renewable energy in Australia, systematically lower its costs, and support and stimulate innovation throughout the value chain, from early research to initial commercial deployment. Attempting to answering a shared need of mines and remote communities for reliable and less volatile costs for power supply, it was expected that a self-sustaining remote power supply market could develop, if renewable energy projects funded by ARENA's RAR program could shine a light on and tackle the structural barriers of governance, supply chains, and access to finance, through pilot projects serving mining and off-grid communities.

| Regional Australia's Renewables program design
The fundamental premise of the design of the RAR program, and offgrid investment for ARENA, was that project economics, subsidy reduction potential, and strong forecast power demand growth constituted a powerful "cost pull," but that an additional "technology push" was needed to demonstrate to end-users that those theoretical cost savings could be realised without sacrificing power quality, reliability, and security. By building a suite of projects across community and industrial energy use, ARENA could retire risk perceptions surrounding the technology and provide evidence of cost savings. The more forgiving cost environment of off-grid power supply meant that renewable energy (particularly solar PV) could be built and derisked, at a time when utility-scale PV was struggling to gain momentum on the Australian grid. Furthermore, by stimulating high renewable energy fraction projects on isolated grids, it was (and is) possible to test renewable energy integration solutions in the off-grid "sandbox," with the findings then available for translation back on the main grid when needed.
The RAR program was designed to provide a total funding of up to AU$400 million (US$384M) for both off-grid projects and fringe-of-grid projects, with 2 separate funding streams focused on renewable energy supply for remote and regional communities (Community and Regional Renewable Energy) and an industry funding stream, aimed at mining and tourism (RAR-Industry). The size of the RAR program funding to ARENA's total funding for projects at the time (AU$3200M; US$3072M) was designed to approximate the ratio of the remote and off-grid power demand to the total power demand in Australia.
To address the structural barriers to the commercial competitiveness of off-grid and fringe-of-grid renewable energy, the RAR program sought to tackle first mover costs and risks, and perceptions of the high cost and unreliability of renewable energy by funding a portfolio of demonstration projects. The program aimed to build 150 MW of total renewable energy capacity, including at least 2 utility-scale plants of ≥10 MW each, and to ensure that projects would be utilised for at least 5 years following commissioning. Sharing of project data and lessons learnt was added to the program, and included in the contractual obligations for successful applicants. By funding a suite of demonstration projects, ARENA also hoped to address strategic "roadblocks," such as risks associated with the integration and management of renewable energy in a hybrid system, and increase skills and capacity associated with renewable hybrid systems in remote areas.
Australian Renewable Energy Agency's funding for the RAR program followed a stage-gate process that is still ongoing: 1. Expression of interest: Applicants were invited to submit a preliminary project proposal.
2. Full application: Following assessment of expression of interests, short-listed projects submitted a full project proposal to ARENA.
This entailed a lengthier and more in-depth assessment process that involved external expert due diligence studies. To better enable communication among stakeholders and to tackle the lack of data capture standardisation, ARENA's RAR Data Specification 19 was developed. While the projects in the RAR portfolio will provide performance data and information to ARENA for subsequent distribution to the market, many more project configurations exist which are not yet captured in the RAR portfolio, and for which additional data will be beneficial. A description of the portfolio of active projects and the technologies used is given in Table 1. PV is the shared generating technology within the portfolio, although it generally needs to be supplemented by other technologies such as wind power or battery storage to achieve the design high annual Renewable Energy Fractions.

| Costs and timing of Regional Australia's Renewables projects
To date, most active projects have been within 10% of their agreedupon cost budgets. Figure 3 illustrates the scaled project costs, the design annual renewable energy fraction, and the number of renew- where net renewable capacity is the DC (or DC equivalent) nameplate net generating (ie, solar and wind) capacity, and excludes storage.
As can be observed from Figure 3, costs for projects vary strongly, Given the above, a more detailed look at the projects can give an indication as to what constitutes a "remote premium."

| Remoteness premium
With the data from the RAR projects' financial models at the time of   Lord Howe Island's application for wind turbine approvals has recently been denied, and the project proponents and ARENA are evaluating which changes are necessary for the project to continue.  Figure 4 were the basis for funding by ARENA to the projects, and for the projects to proceed to securing financing, and signing engineering, procurement, and construction contracts. Nevertheless, contrasting the LSS projects to the RAR project portfolio, a scale and remoteness premium is clear. Given the many variables involved for the different projects, further study will be required to fully quantify the remoteness premium.
On the timing side, projects have-with few exceptions-not been able to complete within the planned, or expected, timeframe. The project duration is calculated from full application to commercial operation. For the projects yet to be completed, the expected completion date is calculated assuming that the project delivery plan is executed as planned with no knock-on delays.  only so far from the local load centre while remaining within economical limits, which thus lead to higher enabling technology needs and costs.
With extended project development timeframes, personnel change becomes a larger risk for all stakeholders. Similarly, significant electricity infrastructure upgrades tend to be few and far between, which carries the risk that the institutional memory of the local stakeholders may lapse. The experience of the RAR portfolio shows that the significant changes to the electricity infrastructure that these projects entail can and do take longer than the continued presence of decision-makers who changed throughout the course of the projects, causing a loss of the context of the project, the decision-making process, and the incentives that drove these. This is compounded by the general lack of experience with renewables integration by local decisionmakers, and the large variety in relevant regulations which often hamper the approvals and project delivery process. In some cases, project developers in the RAR portfolio needed to restate the business case for a project to a new set of decision-makers, after having already committed to it. While this type of risk is common among many infrastructure projects, the long project lead times and relative unfamiliarity of stakeholders and decision-makers with renewable energy means that this risk is magnified for off-grid renewables.
The prolonged negotiations involved in the projects have also resulted in the risk of more stakeholders being involved, each with its own agenda and incentives. In 1 case, a project started with 4 main stakeholders. One of these, the state-owned electricity utility, was Australian and international standards. The rectification work to achieve factory or site acceptance test certification then adds cost and often creates knock-on delays for the project.

| Lessons from the Regional Australia's Renewables portfolio
The key lesson (or rule) which is often forgotten, overlooked, or underestimated by stakeholders for (renewable) electricity supply in remote locations, be they regulators, funding agencies, private sector participants, or end-users of electricity, is that place, people, and communication matter.