Together, taking the next step: using system dynamics modelling to build community renewable energy programmes in Aotearoa New Zealand

ABSTRACT

The Aotearoa New Zealand electricity sector is undergoing a transition to more distributed electricity generation, largely driven by the private sector. Local actors, however, are increasingly interested in playing a role in the energy transition, and addressing their resilience, with community renewable energy (CRE). These projects often have a wide range of impacts, dependent on the community’s motivations, as well as the design and operation of the project. However, there is a lack of tools that can accurately quantify the impacts derived from CRE projects, and guide communities to build a portfolio of projects. This study used system dynamics modelling to analyse the impacts of CRE projects over a 40-year time span. The Energise Ōtaki community group in Aotearoa New Zealand was used as a case study given their already established CRE project and goals to expand their project portfolio. The system dynamics model showed if the community reinvested 80% of the profits from the first project into a second project would provide the greatest long-term benefit for the community in terms of the impact areas important to them. System dynamics modelling proved a useful tool in evaluating the wide range of impacts, and how they interact.

Introduction

What is community energy?

Due to the ambiguity of the term ‘community’, the concept of community energy can has different meanings to different actors, and initial research into the community energy sector attempted to define the boundaries for community energy (CE) projects. One of the most cited articles by Walker and Devine-Wright (2008) attempts to define community energy projects as those that have (a) distributive impacts to respective communities, and (b) have a collaborative development and operational processes. The term ‘community’ can also refer to a geographical position (as in the traditional sense) or refer to common interests shared by a group of people from different areas of regions (Walker 2022). National governments, such as the United States of America (USA), United Kingdom (UK), and the European Union (EU) also have their definitions of what community energy means in their regions (Community Energy England 2024; European Union 2024; Groundswell 2024). The Aotearoa New Zealand government (MBIE 2019) defines CE as energy activities that are ‘managed in an open and participative manner and have local collective benefits and outcomes’ – broadly following the definition used in literature by Walker (2022).

The growth of CE has rapidly increased in the twenty-first century, particularly in Europe (Schwanitz 2023) and state of the sector reviews have begun tracking the growth of CE at regional and national levels. Seyfang et al. (2013) conducted an independent survey of CE projects in the UK, covering one hundred and ninety community energy groups. REScoop (2024) has over 2250 energy cooperatives as a part of their network in Europe, which includes more the 1,500,000 members. The European Commission have also identified over 3700 projects (Caramizaru and Uihlein 2019a). A list of community solar projects and supportive policies in the USA are provided by the National Renewable Energy Laboratory (2024). Further, Groundswell (2024) has identified over 460 low and middle income (LMI) community solar projects in the USA and provides an interactive map and dataset with project details. Community Energy England (2024) has an online map of the projects within their networks. Berka et al. (2020) found 198 CE initiatives in Aotearoa New Zealand and they identified various enablers or barriers to their uptake. This is often done using a variety of tools or analytical frameworks, such as Strategic Niche Management (Seyfang 2014), the Multi-Level Perspective (Ford et al. 2019), and the Social-Ecological System Framework (Bauwens et al. 2016), to develop theories as to the key ingredients to a thriving CE sector within a region. In the Aotearoa New Zealand context, the adoption of CE is dependent on policy measures that are ‘shielding’ (protective of current processes), ‘nurturing’ (supportive of the CE development process) and ‘empowering’ (enabling of adopting and upscaling), and the abilities of national or local governments to co-ordinate these sets of polices determine the ‘window of opportunity’ for CE to thrive (Berka et al. 2020).

The impacts of community energy projects

The impacts of CE projects have also been researched, albeit to a lesser extent. These impacts are far reaching and cross sectoral. Some of the benefits noted by IRENA (2018) include economic sector creation, lower energy prices, social cohesion, and innovation in technologies and business models. Socio-economic benefits through energy ownership, energy education and shared profits that can be reinvested in the community have also been identified by the European Commission (Caramizaru and Uihlein 2019b). These projects can also support an energy system at a technical level by providing energy flexibility, alleviating the need for network upgrades, and reducing electricity transmission losses. These impacts are far reaching and often are determined by the policy environment, governance model and the motivations or objectives of the community organisation (Berka and Creamer 2018; Nolden et al. 2020; Germes et al. 2021; Walker 2022). Furthermore, these impacts are often strongly connected to one another. For example, including many people in the project design and operation (a social impact) can lead to increased environmental impacts in the community as these citizens may become more energy literate and, potentially, save electricity in other areas of their lives.

Community energy can also lead to negative impacts, particularly when a targeted policy does not fully grasp the context-specific social and economic environment. In cases where little state support is provided and market mechanisms are used, the ethics and inclusivity of these projects are often questioned (Catney et al. 2014; Taylor Aiken 2017). In other cases, policies that have looked to stimulate the growth of CE have actually led to a slower shift to a clean energy system (Tews 2018). Projects that should disburse a portion of profits to nearby communities have done so without an understanding of current community development practices (Wlokas et al. 2017). These instances question the efficacy of CRE projects in distributing benefits to community members. Designs of off-grid energy projects also need to grasp social dynamics if they are to ensure the sustainability of the initiative, requiring a holistic and transdisciplinary approach (Brent and Rogers 2010). Project design and implementation therefore play a significant role in shaping the impacts of CRE projects. However, there is little guidance for communities seeking to achieve specific outcomes for the community as to how to develop projects or invest project earnings.

Modelling the impacts of community energy

Academic and grey literature on CE impacts often rely on survey and case data. Impacts are also reported by communities involved in the projects with little information on the methodologies used to collect said data, which leaves room for a bias toward positive impacts. Modelling and forecasting tools that can predict impacts of community renewable energy (CRE)¹ projects are under-used, and predominately focussed on socio – economic regeneration (Berka and Creamer 2018) or to ensure energy demand meets supply (Huang et al. 2015). Wider energy modelling tools also place a focus on technical energy planning and cost optimization (Connolly et al. 2010). These tools and approaches are unable to include the wide range of impacts (social, economic, environmental) that are delivered by CRE projects, and also do not identify the strong links between these impacts. For example, a community may start a CRE project to reduce their carbon emissions whilst also providing extra disposable income to households through grants. Certain households may use this disposable income to purchase cars and their usage could negate any carbon emissions savings gained by the CRE project. In this case, the overall environmental impact (carbon emissions reduced) is dependent on the socio-economic impact (grants to households). These relationships could also show interdependence or independence of certain variables, and in most cases be unique to a project.

Additional to the wide ranging impacts, the success of one CE project often leads to the development of further CE projects. For example, Roberts et al. (2023) describe how iwi communities have grown their portfolio of geothermal energy projects, in alignment with their community values and principles, using the successes of initial CE projects and strategic partnerships. At the same time, grants are often disbursed to communities in the form of health and social grants, education scholarships or stipends for the elderly. Previous research into this question has been done using location-specific social accounting matrices and computer generated equilibrium models (Allan et al. 2011; Phimister and Roberts 2012). These show that local ownership is seen to be crucial to maximise the socio-economic impacts of the communities being analysed. Furthermore, the revenue generated from CRE projects should be reinvested into local capital to maximise socio-economic impacts (Phimister and Roberts 2012). Suitable modeling tools should also be able to assist communities in deciding how they should disburse financial gains from CE projects to maximise long-term community impacts, whilst in alignment with their community motivations and goals.

System Dynamics (SD) was seen to be a suitable choice of modelling the green economy by UNEP (2014) due to its ability to model complex systems, link a wide range of sectors, as well as incorporate feedback loops. Initially developed in the 1950s by Jay Forrester at the Massachusetts Institute of Technology, SD aimed to assess to the successes and failures of large corporations when faced with complex problems (Sterman 2000). Complex systems, like many social systems we see in the world, are not able to be understood by humans and computer software is needed to predict the behaviour of these systems. A system, defined as a collection of components that interact with each other, is not only the sum of its parts but also the interactions between each component (Maani and Cavana 2007). SD models that recognise the high interconnectedness between variables and feedback functions are often used to run ‘what if’ scenarios. For example, a community may start a CRE project to reduce their carbon emissions whilst also providing extra disposable income to households. A household may use this disposable income to purchase a new car, and thereby increase the carbon emissions of the community. The nature of these links between variables is often highly dependent on the case being analysed, and the structure of the model develops theories about the behaviour of the system (UNEP 2014). Instead of adopting general economic theories for the causal relations forming the structure of the model, simulation models propose a theory of their own. System dynamics modelling (SDM) is able to analyse complex systems, which feature uncertainty and feedback loops, and improve these systems by identifying leverage points that weaken undesirable, or enhance desirable feedback loops (Hafner et al. 2021). Furthermore, SDM can incorporate the relationship between sectors (such as energy, agriculture, and so forth) as well include delayed feedback showing how the impacts of one project can affect the planning of another.

This aim of this study is to develop an SDM to assess the impacts of a CRE project and identify ways in which to enhance any positive, and mitigate the negative, impacts. To do so, a suitable system dynamics model will need to perform the following functions:

1. incorporate a range of sectors (economic, social, environmental) with more detailed indicators (such as economic growth and return on investment for the economic sector) over the time span of the project and potentially further,

2. make the links and describe the relationship between the various impacts, some of which may change over time, and

3. model how the performance and design of one CRE project affects the design of future CRE projects within a community.

The research methodology, including the case study selection and model design, follows in section 2. The results of the modelling process are seen in section 3, with a discussion on the suitability of SDM for the forecasting impacts for CRE projects in section 4. Section 5 concludes the paper, and a detailed annex is attached providing more details to the model design.

Research methodology

Research boundaries

Defining of terms and selection of energy activity

For the purposes of this study, the definition of ‘community energy’ used will be similar to that used by the New Zealand Government’s Ministry of Business, Innovation and Energy (MBIE 2019) when referring to community energy projects in this research. This definition is ‘energy activities (such as generation, demand side management, storage, clean transport and energy efficiency) that are managed in an open and participative manner and have local and collective benefits and outcomes’. Community renewable energy (CRE) will refer to only generation from renewable energy sources, such as solar energy, geothermal energy, wind energy and biomass.

As these energy activities would likely lead to different impacts, the research chose to focus on one activity to assess the suitability of SDM. Energy generation was selected as the energy activity, and community renewable energy (CRE) was the focus for the impact framework and the system dynamics modelling.

Case study background information

Energise Ōtaki – a community group, and a registered charitable trust, in the Kāpiti Coast District, New Zealand – was chosen for this study, given the availability of data and the current desire by the community to expand its range of CRE projects. It is also one of the few successful, operational CRE programmes in New Zealand. The organisation’s main goal is to ‘achieve a sustainable and affordable energy future’ (Energise Ōtaki 2018). Areas in which Energise Ōtaki are involved include: renewable energy generation; energy use and conservation; transport; energy recovery; and energy monitoring and research. The group have several projects such as the Warm Up Ōtaki programme that provides grant support for household insulation (Energise Ōtaki 2020). Energise Ōtaki is led by five local trustee members, who all volunteer their time. The trustee members have various backgrounds from business, social sciences, journalism and engineering.

In October 2020, Energise Ōtaki received a NZ$407,000 grant from the Wellington Community Trust to further its work in renewable energy generation. The grant was used to install two solar photovoltaic (PV) systems. The first, a 23 kilo Watt (kW) system at Ōtaki College, supplies clean energy directly to the college. The second project of nearly 107 kW is located at, and provides power to, the Kāpiti Coast District Council’s wastewater treatment plant. The system provides approximately 72% of the wastewater treatment facility’s annual energy usage and excess power is fed back into the grid if not used by the facility. Ōtaki College and Kāpiti Coast District Council pay market rates for the electricity they use, while excess electricity fed into the grid is remunerated at approximately² 10c per kWh by the energy retailer Meridian. This business model – a power purchase agreement with a local organisation with constant energy demand – is also one that could be replicated in other parts of the country. The revenue of the solar operation goes into the Whakahiko Ōtaki – Energise Ōtaki Fund (WŌF) with the purpose of funding community, sustainable energy projects. The WŌF is disbursed on an annual basis and is approximately NZ$23,000 per annum, based on the revenue generated from the solar PV systems. This approach grows the initial wealth invested over time and allows it to cascade down through funded projects into grass roots change and community benefit.

Energise Ōtaki’s goal is to expand its portfolio of solar PV projects and have NZ$100,000 annually going into the WŌF. They have identified five sites with the potential to have solar electricity generated ‘behind the meter’, where power is purchased directly from the solar installation, and revenue is directed into the WŌF through power purchase agreements. These plans include the possibility of Energise Ōtaki accessing a portion of the revenue from WŌF to develop more CRE projects. Even though Energise Ōtaki developed and manage the CRE project, they are still eligible to apply to the WŌF.

Development and validation of the system dynamics model (SDM)

Problem conceptualisation

Energise Ōtaki can use the revenue generated by the current solar installations to invest in new CRE projects, in addition to disbursing funds from the WŌF to the community. Decisions of how much money should be invested into other projects, and at what stage this investment should take place, need to maximise the impacts of importance to the Ōtaki community. Figure 1 conceptualises this problem. The CRE project (illustrated by installed CRE Capacity) has a direct relationship with how much electricity is generated, and how much revenue and profit are generated. These profits can be put towards investing into new CRE projects, and thus increasing CRE capacity over time (shown by the reinforcing loop R). Another option could be to disburse these funds to the community, which would negatively impact the amount invested into CRE projects (shown by the balancing loop B) and is currently the status quo. Both these options will lead to CRE impacts. However, investing in further CRE projects will ensure these impacts accrue over a longer period. Key information that needed to be gathered from the SDM are: (a) how much should be invested, (b) when this investment should take place, and (c) how the investment will be made (loan or directly from the savings from the first project).

Figure 1.

Problem conceptualisation for SDM model.

Apart from the investment decisions, the Energise Ōtaki group identified key areas they wish to address, which were included in the analysis (see Table 1).

Table 1.

List of relevant impacts of the Energise Ōtaki CRE programme.

Variable	Description	Unit
Local economic impact	Local economic impact generated by the CRE project	$/year
CO2 emissions saved	Carbon emissions abated	CO2eq/year
Electricity resilience	Percentage of energy produced in comparison to the energy consumed.	%
Community involvement	Number of local citizens included in the energy transition	People/year

Model design

The design of the Energise Ōtaki CRE Programme Model is based on the systems thinking and system dynamics methods outlined in Maani and Cavana (2007) and Sterman (2000). The full stock and flow model can be seen in Figure 2. The Energise Ōtaki CRE Programme model can be broken down into 6 components, namely:

• Component 1: Project development

• Component 2: Energy generation and revenue

• Component 3: Expenses and disbursements

• Component 4: Project impacts

• Component 5: Energy literacy

• Component 6: CRE 2 investment

Figure 2.

Energise Ōtaki CRE programme model.

A detailed design of each component, along with variable equations, can be found in Annex 1 of the supplementary material. The blocks in the SFD represent stocks and exogenous variables are indicated with circles. Vensim software (Smart 2021) was used to design the model. A time step of 0.0125 years was used with a Euler Integration method.

The solar PV project was in its second year of operation when data was collected, providing historical data points for the model. A set of engagements were held, and a survey designed, to obtain the data and guidance from the Energise Ōtaki community members. These included:

1. An online call to propose the research, explain the process and the work done and agreement of Energise Otaki to participate.

2. Emails and online discussions of the survey for data collection.

3. On-site visit to the solar farm, followed by a discussion on the system with the community group.

4. On site visit to community group offices to present the model, gather further feedback and a general discussion of the model’s suitability.

Feedback loops

Reinforcing feedback loops: CRE revenue

Three reinforcing feedback loops were identified in the model, shown in Figure 3. All three feedback loops start at the CRE Capacity variable and move to the Electricity Produced By CRE Per Year. Loop R1 involves the revenue being generated by electricity fed into the grid, and is dependent on the Electricity Fed Into The Grid and the Average Feed-In-Tariff. Similarly, loop R2 involves the electricity revenue from KCDC. This is dependent on the Electricity Used By KCDC and the Average Electricity Tariff Paid By KCDC. Loop R3 involves the electricity revenue earned by the selling of carbon credits (if this option is selected by Energise Ōtaki). This involves the Electricity Produced By The CRE Per Year and the Carbon Credit Price. All three of these reinforcing loops then feed into the Whakahiko Ōtaki–Energise Ōtaki Fund stock and along the same path back to the CRE Capacity, through the CRE Project Investment Fund Stock and the CRE 2 Capacity variable. The higher the revenue from (a) the electricity fed into the grid, (b) the electricity used by KCDC, and (c) selling of carbon credits, the higher the expenditure on capital leading to an increase in the CRE Capacity. There is also a direct relationship between the CRE capacity and these three revenue streams – creating the reinforcing loops. There is a delay in the revenue leading to an increase in CRE capacity, as the second CRE project will only be commissioned after CRE 2 Investment Period. This delay is affected by the CRE 1 Lifespan shadow variable in the CRE Project Investment Fund stock. Only once CRE 1 Lifespan reaches CRE 2 Investment Period, will the funds in the CRE Project Investment Fund be spend on a second CRE 2 project.

Figure 3.

CRE revenue feedback loop.

Balancing feedback loops: CRE expenses

Figure 4 shows the balancing feedback loops seen in the Energise Ōtaki CRE Programme model. This figure has been adapted slightly to better show the feedback, and shadow variables are used in the actual model. Both feedback loops start at the CRE Capacity variable. Loop B1 involves the insurance cost for the CRE project, and is determined by the CRE Capacity and the Insurance Cost Ratio. Loop B2 involves the operations and maintenance expenses for the CRE project, and is determined by the CRE Capacity and the Operations And Maintenance Cost Ratio. Both these feedback loops then flow into the Whakahiko Ōtaki – Energise Ōtaki Fund stock and, similar to the reinforcing loops, impact the CRE 2 Capacity variable. The higher the insurance, operations and maintenance costs, the less funds are available for the second CRE project, and a smaller CRE 2 Capacity. A lower CRE 2 Capacity will lead to lower insurance, operations and maintenance costs, thus creating a balancing feedback loop. A delay in this feedback is experienced, as explained in the reinforcing loops above.

Figure 4.

CRE expenses feedback loop.

Model validation

When designing a SDM, it is crucial to have a reflective testing process and aim to identify potential flaws in the model construction (Sterman 2000). A variety of tests can be used to refine a model, each with its own purpose. For the Energise Ōtaki CRE Programme Model, five main tests were used to help identify weaknesses in the model (detailed in Sterman (2000), Chapter 21) and are described in Table 2.

Table 2.

Accuracy tests conducted on the Energise Ōtaki CRE programme model.

Test type	Purpose	Tests performed
(1)Boundary testing	Assess the behaviour of the model if the boundary assumptions are changed significantly	Testing project financing Testing of business model Varying of CRE capacity Varying of time span
(2)Structural assessment	To ensure the system behaves consistently with knowledge of the real system and relevant to its purpose	Design and comparison to behaviour over time graphs Regular simulations through model design
(3)Sensitivity analysis	Analyse if the results of the model change significantly when exogenous variables or estimated data is varied	Sensitivity to capacity factor Sensitivity to local economic impact multiplier Sensitivity to indirect carbon emissions abatement factor Sensitivity to grant size Sensitivity to loan parameters
(4)Dimensional assessments	Assess whether equations and units are aligned during calculations and have real world meaning	Inspection of equations and units of variables through chosen software and manual review
(5)Relevance and suitability assessment	Assess whether the model is of use to the Energise Ōtaki members	Presentation of modelling approach and results to the Energise Ōtaki Group

The results of these tests, potential model weaknesses they exposed, as well as the solutions to solve these weaknesses are described in detail in the Annex of the supplementary material.

Scenario analysis

Three primary variables needed to be tested through the scenarios, namely: (a) how much should be invested into the second CRE project, (b) when this investment should take place, and (c) should the investment come from savings of the initial project or a loan. The first variable is determined by how much of the revenue from the first project is reinvested into a second CRE project, as opposed to being distributed to the community – termed the CRE 2 Investment Ratio. The second variable is the period after the first CRE project is commissioned and is termed the CRE 2 Investment Period. The last primary variable to analyse is the effect of taking a loan for a second CRE project. Table 3 provides a summary of the scenarios and their descriptions.

Table 3.

List of scenarios to be tested with the Energise Ōtaki CRE programme model.

Scenario	Description
Early re-investment	The early re-investment will save of 50% of the annual revenue generated from CRE project 1, and use this to develop a second CRE project 10 years after CRE 1 project is operational.
Increased early re-investment	This scenario will save of 80% of the annual revenue generated from CRE project 1, and use this to develop a second CRE project 10 years after CRE 1 project is operational
Late re-investment	This scenario will save of 50% of the annual revenue generated from CRE project 1, and use this to develop a second CRE project 25 years after CRE 1 project is operational.
Early re-investment with external finance	This scenario will save of 50% of the annual revenue generated from CRE project 1, and use this to develop a second CRE project 10 years after CRE 1 project is operational. A loan will also be taken out to increase the amount available for the second CRE project. The loan amount would increase the total investment into the CRE 2 project to an amount similar to the Increased early re-investment scenario.

The timespan for the scenarios was set to 40 years to analyse the effects of multiple projects. This was chosen as the Early Re-investment scenario would end at this point. Residual impacts will occur for the Late re-investment scenario after 40 years, given the estimated lifespan of a CRE project is 25 years. However, a planning timeframe of longer than 40 years was seen to be unrealistic for a community organisation. Similar parameters for the second CRE project (such as feed-in-tariffs, operations and maintenance expenses, rent and self-consumption ratio) were used for the results of the simulations, although the model can accommodate any changes to these.

Table 4 shows the scenarios run for the various investment options for the second CRE project, and the respective variables adjusted. The capacity of the second CRE project is a function of how much is saved from the first project and the amount of money taken as a loan. The resulting power capacity of the second CRE 2 project, given the various investment parameters, are also shown in Table 4. Figures 5–8 show the results of the scenarios and Table 5 provides the results of each scenario at year 40.

Table 4.

Scenarios run through Energise Ōtaki CRE programme model.

	CRE Investment time	CRE investment ratio	Loan Parameters			Capacity of CRE 2 Project (KWp)
			Loan amount (NZ$)	Loan Interest rate (%)	Loan Term (years)
Early re-investment	10 Years after CRE 1	50%	0	0	0	38
Increased early re-investment	10 Years after CRE 1	80%	0	0	0	60
Late re-investment	25 years after CRE 1	50%	0	0	0	95
Early re-investment with external finance	10 Years after CRE 1	50%	65,000	10%	25	60

Figure 5.

Local economic impact for scenarios of the Energise Ōtaki CRE programme … 

Figure 6.

Total carbon abated for scenarios of the Energise Ōtaki CRE programme.

Figure 7.

Citizens included in the energy transition for scenarios of the Energise Ōtaki CRE programme.

Figure 8.

Electricity resilience for scenarios of the Energise Ōtaki CRE programme.

Table 5.

Cumulative results of the key variables after the 40-year period Energise Ōtaki CRE Programme.

	Early re-investment	Increased early re-investment	Late re-investment	Early re-investment with external finance
Local economic development ($000)	1,245	1,346	1,057	1,172
Total carbon abated (tons CO2eq)	730	799	632	718
Citizens included in the energy transition	943	996	726	812
Average electricity resilience (%)	0.84%	0.96%	0.87%	0.96%

Results

Local economic impact

The results show that increasing the CRE 2 Investment Ratio provides the highest long term local economic impact (as seen in Figure 5), leading to over $1.3 million over the 40-year time span – a more than 300% increase from the initial grant. Starting the second CRE project after the first (as is the case with the Late Re-investment scenario) will lead to the lowest local economic impact out of the scenarios. An increase in the CRE 2 Investment Ratio leads to a higher local economic impact. However, this is a slight change over the 40-year period.³ As expected, taking a loan to increase the capital expenditure needed for a second project sees a reduction in the local economic impact, in comparison to the Early Re-investment scenario. This is because interest rate payments reduce the amount distributed to the community. Figure 5 shows the local economic impact over time for each scenario.

Carbon emissions abated

The carbon emissions abated graph seen in Figure 6 follows a similar pattern to the local economic impact graph seen in Figure 5. The higher the CRE 2 Investment Ratio, the less attractive the results over the first half of the period. However, more favourable results are seen in the long term. An increase in CRE 2 Investment Ratio leads to more carbon abated⁴ than the Early Re-investment scenario, yet this effect is less substantial than increasing the CRE 2 Investment Period. Almost 800 tons of carbon is estimated to be saved with a CRE 2 Investment Ratio of 80%.

Citizens included in the energy transition

Figure 7, showing the citizens included in the energy transition (through being awarded grants), has similar behaviours to that of the Local Economic Impact and Carbon Emissions Abated results. The Increased Early Re-investment scenario sees a reduction in citizens awarded grants in the short term. However, this accelerates after year 13 and eventually yields more citizens being awarded grants than the Early Re-investment scenario. Awarding almost 1000 people grants, is quite substantial. Given the potential energy literacy benefits from CRE, the number of citizens awarded grants should be a focus, and could increase the impacts of all the other variables analysed in this model.

Electricity resilience

The electricity resilience of Ōtaki is impacted minimally through the CRE projects, as seen by the scale in Figure 8. At its peak, the Programme Base Case scenario increases the electricity resilience of the community to 1%, and gives an average electricity resilience of 0.84% over the period analysed. Increasing the CRE Investment Ratio, or taking a loan for a portion of the second CRE investment costs, increases the average electricity resilience as more money is available for a CRE system with larger power capacity. Delaying the construction of the second CRE project reduces the peak electricity resilience of the CRE programme and gave the lowest average electricity resilience over the period.

Discussion

Summary of simulation results

The results from the simulations show that a second CRE project, with the revenue generated from the first project, will give positive results on the key variables analysed over the 40-year timespan, with the ideal scenario being a larger CRE Investment Ratio of 80%. A significant factor is the indirect carbon abated (i.e. from sustainable energy projects initiated by community members using seed funding from the CRE project’s profits) through the distribution model Energise Ōtaki employs. Using the estimates given, these indirect impacts outweigh the direct benefits from the CRE project itself. This shows the compounding, positive environmental impacts CRE projects can have.

More broadly, it is significant that, in this scenario, the initial grant of $407,000 can leverage over $1.3million in economic benefits, award almost 1000 community members grants, and abate over 800 tons of carbon over the 40-year time span, should a CRE 2 Investment Ratio of 80% be followed. These data would be extremely valuable for the community members when deciding how to allocate profits generated from the CRE project. Furthermore, it can also be very useful when trying to attract external funding from other sources.

The outcomes of the research are similar to those seen by and Phimister and Roberts (2012), in showing that revenue generated from CRE projects should be reinvested into local infrastructure, more specifically another CRE project, to maximise the impacts for the community, as opposed to financial distributions directly to community members. Although it may be easy to generalise from these results that further reinvestment is a better option, a deeper look is needed. This research showed that carbon emissions saved also increased when reinvestment into further infrastructure is prioritised, but this is dependent on the type of infrastructure being developed. In the case of Energise Ōtaki, this was a second CRE project, however this may not be so for other communities reinvesting in other projects. Given that more citizens will receive grants over the 40-year period when more project revenue is reinvested into other CRE projects is also crucial for wider impacts associated with CRE. As noted in the introduction, the participation in, and benefaction of, citizens in CRE projects are reported to improve their energy literacy⁵, and help them make behavioural changes to reduce their energy usage and carbon footprint. Furthermore, including communities within projects is also posited to increase acceptance of renewable energy projects situated close to community areas, and avoid delays in transitioning to a clean energy system. Reinvestment into further CRE projects by community members, whilst still ensuring community involvement, should be maximised to obtain overall community impacts. More broadly, this research, shows that CRE can be a mechanism for wider, low carbon community development, and planning and investing in CRE projects can ensure the positive impacts of a global energy transition reach local communities.

Suitability of SDM for CRE programmes

SDM was able to fulfil the three criteria set out to address the research objective. The model was able to incorporate variables from a range of sectors, more specifically social, economic and environmental sectors. The model was also able to link the various indicators and show the relationship between them. Impact modelling tools used for CRE projects have previously been limited to either socio-economic (Berka and Creamer 2018) or technical in nature (Huang et al. 2015), which does not match the wide range of impacts that can emanate from CRE projects. SDM can be used to fill this gap, and provides a holistic impact modelling approach A good illustration of this is looking how taking a loan for a second CE project, with specific loan parameters, can affect the number of citizens receiving grants and the number of carbon emissions abated by the community.

Given the data requirements an SDM is perhaps more useful in the pre-feasibility stage of project development. Once a project has been conceptualised, and initial scoping studies conducted, project level system dynamics modelling could add additional information to potential project impacts, given certain funding structures and implementation models. Once an SDM model is created community members may be able to use the model to forecast the impacts of a project or identify levers that maximise positive impacts and mitigate negative ones. The uniqueness of each CRE project makes it more suitable to design project specific models as opposed to a generic model that can be adapted to each project. A standardised list of indicators, or impact areas relevant to the community, and methodologies to calculate these impacts, can provide comparable results between models. Knowing a system dynamics modelling approach will be taken to analyse project impacts from the start of the development process may also help alleviate data collection challenges, as relevant data needed can be pinpointed from the project conceptualisation stage.

Finally, a link between the performance of the first CRE project was drawn to the establishment of the second CRE project, which is often seen in CRE projects within New Zealand (Roberts et al. 2023). Using an SDM can provide communities with data that can inform their decision on how best to plan for a further CRE projects. The outputs of the CRE SDM model could provide further information for financiers that may be willing to support clean energy, community-led projects – such as the many trusts linked to the electricity distribution industry in New Zealand or impact investors. As previously stated, there is a lack of understanding of the wide range of impacts and their magnitude from CRE projects, creating difficulties to lobby the different tiers of government who could provide an enabling environment for these projects.

Feedback from Energise Ōtaki

The Energise Ōtaki CRE Programme Model was discussed with the Energise Ōtaki team, and in general the members were impressed with the modelling approach taken. This modelling approach would be useful in guiding them to make decisions about the design and implementation CRE projects. One comment was that the ‘human’ element to the CRE impacts was said to be missing, referring to the relationships created, and knowledge shared within the community in the implementation of the CRE project. The use of Likert Scales and surveying community members may provide quantitative information that could be an addition to this modelling approach, although it may not suitably capture this impact. Qualitative information, through case studies and storytelling, and quantitative information is best used together when accurately describing the CRE impacts. Quantitative data from SDM models is better used to describe the scope or spread of the impacts (e.g. number of citizens with energy bills savings) and qualitative impacts is better used to describe the depth of said impacts (e.g. what that energy saving means to a low-income household). Given this, a forum for storytelling or qualitative information needs to go hand-in-hand with collection and development of quantitative impact data. However, proxy variables, such as Citizens included in the energy transition can be connected to some of these qualitative indicators.

Conclusion

The research showed that an SDM is a suitable modelling tool for community groups looking to develop CRE programmes. The tool was able to describe the impacts emanating from CRE projects, link these impacts, as well as simulate how a change on one can affect the entire system. Qualitative data could be used to deepen the description of these impacts, and better convey what the results mean to community members. The feedback loop function of an SDM was crucial in showing how decisions from one CRE project affect the investment available for future CRE projects, and the long-term impact on the community. The Energise Ōtaki CRE programme model was able to guide investment decisions for future projects to the Energise Ōtaki group, and was well received by the steering committee.

Collecting accurate data for SDMs remains a challenge for communities, and better data will improve the modelling results. Stock and flow diagrams designed using an SDM can visually depict the relations between variables well, making it easier to describe the individual components of the model and showing the various feedback loops. However, it does take some capability to build a model, and it is highly unlikely community members involved in the design and development of the projects will have the capacity to construct an SDM. The Energise Ōtaki group comprised of highly skilled, technical members and were able to understand the modelling approach and results. However, this may not be the case for other communities. External stakeholders or agencies that support communities in developing CRE projects and programmes may be better suited to perform this function. The scarcity of funding sources, and uncertainty in future project parameters, especially on feed in tariffs, also remains a challenge when using SDM to forecast CRE project impacts to influence design.

This research filled a gap in the CRE research field by using SDM to forecast impacts of CRE project. More broadly, it attempts to use modelling tools that can look at a wide range of impacts often emanating from CRE projects. Future research could use system dynamics modelling to understand the impacts of projects with different governance or implementation models, as these have been stated to lead to different impacts. Long term analyses of CRE project impacts would also be useful in validating the results forecasted by models developed for CRE projects. These analyses could include qualitative components and obtain data from wider communities on how they have been impacted by CRE projects.

Funding Statement

This research was supported by the Victoria University of Wellington Scholarship Programme.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplemental Material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/03036758.2024.2385082.