Heat Networks Delivery Models

7. Evaluation

This section provides a summary of the evaluation results, highlighting the main advantages and disadvantages of each delivery model (as described in section 6) against the attributes (as described in section 4).

Appendix B contains evaluation scores for each model, based on the following 0 – 3 / ‘RAYG rating’:

Evaluation Score	Score description
0	Potentially negative impact in relation to this attribute; fails to meet attribute at all.
1	Is neutral in relation to this attribute; neither benefit nor negative impact
2	Performs well against this attribute
3	Performs very well against this attribute.

Table 1 below contains the total scores for each model (against each attribute) and a provisional ranking of models according to these scores. Note that for the purpose of the evaluation exercise, each attribute was given equal weighting.

We have not provided a score against ‘Balance Sheet Treatment’, but instead provided commentary on which organisation’s balance sheet projects delivered under the model would be classified to (e.g., private sector, central Government or local authority).

Our evaluation findings, including the scores, are the output of a series of evaluation meetings, in which commentary and scores developed separately by evaluators were moderated. The evaluation was informed by our own knowledge of the delivery models and the heat network market, and by the feedback received through the stakeholder engagement undertaken in preparing this paper.

Table 1 - Evaluation summary

Ranking: 1

Model Description: Regional ESCo (DM10) – local authorities and other public bodies (e.g., NHS, universities / colleges) would come together on a regional basis and jointly procure a delivery partner for each region (similar to Hub model). The local public partners would use the delivery partner to scope projects and deliver according to pre-defined contracting structures.

Model type: New HN delivery model

Score: 22

Ranking: 2

Model Description: Local authority led joint venture (DM4) – a local authority procures a partner and forms a JV to serve an initial project (including one or more local authority anchor loads) and potentially additional projects and/or other energy projects within the local authority area.

Model type: Existing HN delivery model / limited examples

Score: 19

Ranking: 3

Model Description: Local authority led delivery, with Scottish Government stake (DM9) – local authorities would lead development and delivery of projects, with Scottish Government / central support and co-investment (which may be in addition to an element of grant). Scottish Government would have part-ownership of schemes, but with potential for sale / transfer of government stake once scheme is established.

Model type: New HN delivery model

Score: 18

Ranking: 4

Model Description: Centrally led delivery (DM8) – Scottish Government would take the lead on development and delivery of projects (without need for local authorities to lead development). Scottish Government would have initial ownership / part-ownership of schemes (alongside private sector) but with potential for sale / transfer of government stake once scheme is established.

Model type: New HN delivery model

Score: 17

Ranking: 4

Model Description: Regulated Asset Base (DM12) – a private sector ownership model in which heat network assets are constructed, owned and operated by a monopoly supplier on a long-term basis. Investment plans, operating performance and returns (which are capped) are subject to regulatory oversight.

Model type: New HN delivery model

Score: 17

Ranking: 6

Model Description: Service concession (DM2) – a heat network owned and operated by the private sector under a long-term service concession tendered by a public body, where the public sector offers anchor loads, and concessionaire takes demand risk.

Model type: Existing / well-established HN delivery model

Score: 15

Ranking: 7

Model Description: Public sector (non-Scottish Government) in-house delivery (DM1) – a heat network is wholly owned and operated by public body (either directly, or via a wholly-owned arm’s length entity), usually based on self-supply arrangements (e.g., local authority buildings, or a public sector campus).

Model type: Existing / well-established HN delivery model

Score: 11

Ranking: 8

Model Description: Merchant model (DM7) – a private sector heat network operator contracts with off-takers to supply existing buildings, without having either being appointed by a private sector landowner / developer in connection with a particular development site, or having followed a public procurement exercise.

Model type: Existing HN delivery model / limited examples

Score: 10

Ranking: 9

Model Description: Third party ESCo (DM3) – a heat network owned and operated by a private sector third-party ESCo appointed by private sector land owner / developer, generally to serve new development.

Model type: Existing / well-established HN delivery model

Score: 9

Ranking: 10

Model Description: Unbundled model ( DM6) – a family of models involving separate ownership of generation, transmission / distribution and supply assets, e.g., where heat generators contract directly with customers and pay a use-of-system charge to the owner of heat transmission / distribution infrastructure.

Model type: Existing HN delivery model / limited examples

Score: 8

Ranking: 11

Model Description: Community led project (DM5) – a community leads heat network development and owns the network, subcontracts O&M, supplies buildings within community.

Model type: Existing HN delivery model / limited examples

Score: 7

Ranking: 12

Model Description: Public Private Partnership (PPP) (DM11)- a heat network is operated by private sector under a long-term contract tendered by a public body, where the public sector retains the majority of demand risk, but availability risk lies with the PPP contractor.

Model type: New HN delivery model

Score: 6

It is important to note that the ‘ranking’ reflects the scoring methodology in which each attribute has been equally weighted. However, we consider that the above ranking broadly reflects the most important priorities, i.e. potential for private sector investment, skills & capacity, ease of deployment (alignment with forthcoming regulation; need for further legislation/new bodies) and simplifying delivery (procurement and dealing with stakeholders).

The following points should be kept in mind in relation to this evaluation section:

1. This section should be read in conjunction with the scores in Appendix B. Although scores have been moderated, an element of subjectivity is implicit, and in many cases good arguments can be made for an increase of decrease in an individual score against an attribute by 1 mark.

2. Some of the models (and enabling structures) contain numerous components which operate differently depending on how they are combined or layered. In particular, newer or more innovative models tend to be more complex, and incorporate within them components of other existing models. These components, and how they are deployed, could be flexed to support different outcomes. However, in order to contrast and evaluate different interventions, we have had to make assumptions about how each proposed ‘new’ model would operate in practice. The assumptions in relation to each model are set out in the model’s detailed description, which is contained in section 6. Key assumptions which significantly influence scores but could (in practice) be adjusted are also referenced in the evaluation text below.

7.1. Evaluation of DM1: Public sector (non-Scottish Government) in-house delivery

In this model, the heat network is wholly owned and operated by a public body (either directly, or via a wholly owned arm’s length entity), and is usually based on self-supply arrangements (e.g., local authority buildings, or a public sector campus). This is currently the most common model for heat networks in the UK, and these projects are generally led by local authorities. Therefore, the evaluation below assumes local authorities would be the lead body. A full description of the characteristics of this model is set out in section 6.

As an existing model using established corporate structures and generally involving only one organisation (apart from sub-contractors), this model is relatively easy to deploy. Should heat network regulation fall to local authorities in the future (there is scope for this under the Heat Networks (Scotland) Act 2021), then there is some risk of self-regulation.

This model scores badly for potential for private investment because, by its nature, it involves only public sector investment. Although local authorities may access private finance to provide capital for projects, we have only considered this to be ‘private investment’ if the lender has no recourse beyond the arms-length body/SPV (i.e., the local authority, as ‘parent company’ to the SPV, does not provide a guarantee to the lender). In practice, non-recourse debt is unusual, and projects therefore generally remain entirely on the public sector sponsor’s balance sheet. This is more of an issue for projects led by central bodies such as colleges and NHS Boards, as balance sheet treatment of those projects is likely to have implications for Scottish Government capital budgets (and the potential to impact spending on other priorities).

Although there are numerous examples of successful heat networks using this model, increasing its prevalence may not contribute much to simplifying delivery at pace and scale. Heat networks in general are complex to deliver, and unless there is a successful track record of delivery within an authority, deploying new projects on this basis can be difficult in practice. For a local authority establishing new networks, the process is resource intensive and can often get caught up in changing internal politics and priorities. It also requires an ongoing capacity commitment in order to for projects to be run successfully once established. As a result of all these factors, projects tend to be small scale.

Although authorities that own heat networks develop in-house skills and capacity, experience suggests that this is often difficult to retain. At present, under current resourcing structures, local authorities do not have the resource to dedicate to delivering these kinds of project on a large scale. Even in places where decarbonisation of heat is a priority, resourcing must compete with other (statutory) obligations. Through our engagement with stakeholders, we had consistent feedback that local authorities do not have the resources required to drive forward heat network delivery at the pace required, if they are expected to lead on delivery (or even project identification). Even those local authorities with clearer plans for in-house heat network delivery recognised that they had capacity constraints, and that the level of effort or interest they were committing was unusual, and a result of unique political support in their area.

This model scores well in terms of its ability to contribute to wider policy goals. Retaining public ownership allows wider policies to be promoted and potentially prioritised over profit generation, although it is essential that projects are still run on a commercial basis, to ensure they are adequately resourced and can continue to operate effectively in the longer term. The relatively small scale of 100% public sector owned projects could limit the potential for wider policy benefits.

This model is slightly less sensitive than some others to demand risk, because the project anchor loads are often owned and controlled by the promoting authority, who can therefore manage connection risk. However, projects remain sensitive to demand risk for buildings not within the authority’s control. The appetite of public bodies to accept this risk tends to be low, with projects scaled accordingly.

Although this model is delivering projects in Scotland and therefore supporting development of the market, we do not see it making a significant contribution to the transition to a self-sustaining market. Experience suggests these kind of projects still require significant amounts of grant, and are not offering opportunities for private investment or innovation. As the schemes delivered are often smaller and opportunistic in nature, without clear commitments to longer term growth (or the resources to identify expansion opportunities), supply chain capacity is unlikely to be significantly boosted by a proliferation of this type of project.

As noted above, projects under this model tend to be relatively small scale, as they are challenging to resource, even where the majority of the buildings to be connected are owned by the public authority leading the project. For this reason, it is not a model which can easily be replicated across different geographies, although there is a growing amount of ‘best practice’ knowledge which can be captured and shared.

Expanding publicly owned projects, particularly beyond the estate of the lead organisation, is challenging for most authorities. Authorities are often unwilling or unable to resource the operation of heat networks in a way that facilitates their expansion, in part because the supply of heat to third parties is not considered ‘core business’. For this reason, proliferation of this model would not be expected to support interconnection and expansion of networks, despite the high degree of control which is offered by single, public sector ownership, which makes interconnection and expansion feasible in theory.

Public sector lead organisations may be better placed take a longer-term view of demand and may be able to identify opportunities to invest ahead of need that are not purely profit driven (e.g., in order to accelerate decarbonisation of the built estate). However, such an approach requires the sponsoring authority to carry greater risk. In practice, this kind of investment is usually only achieved when supported by grant subsidy (e.g., capital grants for future proofing), and the only to a limited degree.

By definition, these projects do not have private investment and would therefore appear on the balance sheet of the sponsoring authority. Projects led by central bodies such as colleges and NHS Boards are likely to have implications for Scottish Government capital budgets (i.e., the potential to impact spending on other priorities).

7.2. Evaluation of DM2: Service concession

Under a service concession, a heat network is owned and operated by private sector under a long-term service contract tendered by a public body following a competitive procurement process. The public sector offers anchor loads, and the concessionaire takes demand risk. A full description of the characteristics of this model is set out in section 6.

As with local authority led schemes, there are numerous successful examples of service concessions for heat networks in the UK, and so the model scores well for ease of deployment. The concession model is broadly compatible with forthcoming regulation under the Heat Networks (Scotland) Act 2021, although consideration needs to be given as to how commercial tendering processes can best align with/feed into Heat Network Zones and the subsequent issuing of permits. There are some additional complexities when tendering for a concession compared with (say) a Design & Build contract. In particular, the long-term nature of concessions means that it very important to define and incorporate into the contract the right outcomes at the outset, as it can be complex and expensive to make changes subsequently. This can be challenging, given that unforeseen changes are inevitable over typical concession periods (25+ years). There is, however, a growing body of experience of concessions, from which best practice can be drawn.

The concession model scores well in terms of potential for private investment - the model is predicated upon private investment. The structure can be attractive to the private sector because of the extended concession term (generally 25-40 years for heat networks) coupled with a clear exit strategy (the heat network assets generally revert to the public authority at the end of the term, although some contractors indicated during stakeholder engagement that they would prefer to retain ownership of assets beyond that point.) A higher rate of return is generally sought for this model by investors, when compared with models where the investment is provided by, and/or greater risk is shared with, the public sector. At present, this tends to require higher levels of subsidy to support the investment, and also means that this model is only really suitable for more profitable opportunities. The traditional concession model generally only delivers a single investment, rather than offering an entry-point for ongoing investment through multiple subsequent projects. It is therefore less efficient at deploying private investment than some other options.

Greater deployment of the concession model could help to simplify delivery, but changes would be required to make the process more efficient. Although in theory it is possible to aggregate multiple projects and tender them under a single concession, this requires more up-front work by the authority, which will generally prefer to proceed with the project it has ready at the time. The resulting ‘One Procurement, One Project’ model that we see in use today for concessions is consequently slow and very costly for authorities and developers alike (particularly given that reference designs developed during by the public sector sponsor during the Feasibility and Outline Business Case (OBC) stage are often not ultimately used by the concessionaire). Bid costs and public sector development costs are ultimately paid for by customers and/or taxpayers.

Some newer procurement models, including the ‘Joint Development Agreement’ option discussed in section 6.4 (Enabling Mechanisms), could potentially improve the process by appointing a development partner / prospective concessionaire earlier in the process. We received consistent feedback from developer stakeholders (and, to a slightly lesser extent, investors) that current procurement practices are unworkable, and that bringing in partners earlier (before concept designs are developed) would bring significant benefits and efficiencies. A two-stage procurement process such as the Joint Development Agreement could be one way to deliver this, providing the benefits of such an approach were demonstrably greater than the risks (this has not been tested). The approach could also be applied to other models. Given that concessions are only suited to larger projects with higher potential returns, focussing on this model alone is not likely to increase delivery substantially.

As concessions generally attach to larger projects, they can bring a boost to local supply chains, supporting the development of skills and capacity. Once projects have been procured, the private sector partner can bring their own capacity and expertise, and may be more efficient at both drawing on experience from, and sharing knowledge across, its wider business, giving an opportunity for the skills acquired on one project to be transferred to others. The private sector concessionaire should be well placed to recruit and retain skilled staff, with associated benefits to external advisers. Although the procuring authority will develop skills in procurement and must continue to manage the contract throughout its term, the concession model does not support significant development by local authorities of internal expertise in heat networks.

This model does little to contribute to wider policy goals. Although projects are often large and long term, meaning that any positive contributions the project makes will have a greater impact, securing those outcomes over the long term via an initial contract negotiation is challenging. It is difficult to draft contract conditions in such a way that allows them to be sufficiently flexible over time, without such flexibility being perceived by bidders as a risk. However, as experience grows in this area, there should be an increasing number of examples to draw on which may help authorities secure best value.

This model does not mitigate and is not less sensitive to demand risk. The risk moves from the authority to the concessionaire to varying degrees (depending on the detail of the contract), and is priced accordingly by the concessionaire. The procuring authority can offer anchor loads at the outset to reduce the demand risk to the concessionaire, but that approach is not unique to this model.

Proliferation of this model would help the transition to a self-sustaining market. Concessions support larger projects that cannot generally be delivered by public authorities in-house. Larger projects support larger scale investment, which should, in principle, drive costs down, including cost of capital. Although there are different investors interested in investing in projects of varying sizes, our stakeholder engagement indicated a preference for larger-scale opportunities, i.e., minimum investments in the range of £20–50m. This is particularly true if institutional investors are to be attracted into the market, for whom due diligence costs can only be justified for larger opportunities. The main downsides of the concession model in relation to this attribute are that a) it may not offer a ‘gateway’ to multiple projects, so the potential for a lasting benefit to supply chains and investment opportunity is more limited; and b) any initial capital investment is still likely to be reliant on capital grants in order to meet the private sector’s investment criteria. There is no indication that concessions will be able to move away from grants in the near future, though the same is likely true for most delivery models.

The concession model does not particularly support replicability across different geographies, because it requires projects of significant scale in order to make the investment viable, meaning it is only really suited to larger urban areas.

The concessionaire, for whom heat networks are core business, will be naturally incentivised to identify (profitable) opportunities for interconnection and expansion, and can be expected to resource the project accordingly. However, the model is entirely commercially driven, so it is likely that only profitable and commercial growth opportunities would be pursued, unless the concessionaire was contractually obliged to do otherwise (which would be complex and expensive to establish within the contract). Once let, implementing change in contracts can also be complex and expensive, so may not lend itself to exploiting interconnection opportunities.

Any investment ahead of demand under the concession model would either need to be envisioned, required and paid for by the procuring authority at the outset (e.g., by competing the subsidy to be provided to support a fully future-proofed scheme) or be in the commercial interests of the concessionaire, who may oversize pipes if confident in being able to secure additional customers to make a return on its investment over the concession period. The private sector will generally only accept limited demand risk, and so investment ahead of need may need to be funded by the public sector (or paid for by consumers).

Balance sheet treatment will depend on the level of risk transfer achieved, and any control that the project sponsor can exert, under the terms of concession agreement.

7.3. Evaluation of DM3: Third Party ESCo

In this model, the heat network is owned and operated by a private sector ESCo appointed by a private sector landowner or developer, generally to serve new development. This model has been used predominantly in London, where strict planning conditions require heat networks for large-scale new development. We have assumed this model is limited (at least initially) to the site of any new development. This model is therefore distinct from DM7: Merchant Model, which assumes more of a prospecting approach, aiming to connect new and existing anchor loads across an area. A full description of the characteristics of both of these models is set out in section 6.

This model uses existing corporate structures, does not require a procurement or state involvement (other than planning and administering any regulation). Like most models, consideration will need to be given about how its proliferation would interact with exclusively permitted zones under the Heat Networks (Scotland) Act 2021. Overall, this model scores well for ease of deployment.

This model scores reasonably well for potential for private investment. Privately-owned structures are easy to transfer and exit from, without any state interest or public ownership to complicate investment, purchase or sale. However, investments are likely to be on a project by project basis, mostly likely estate or developments owned by one landowner or body, meaning schemes may not be large enough to attract certain types of capital or support significant investment (see below).

Although working well in specific circumstances, this model would not contribute significantly to simplifying delivery overall. Due to the lack of public sector involvement, this model avoids procurement and so can be delivered quickly. However, generally speaking it is suited for new development only, and has limited potential for retrofit or connection of existing anchor loads: there is less potential for this model in larger areas. Projects are generally delivered on a site-by-site basis, but can leverage relationships between land owners, land developers and heat network developers, replicating strategies that have worked for previous projects.

Although this kind of relationship building can foster skills and capacity in the private sector, a large number of relatively small schemes under this model may draw on private sector skills in an inefficient manner. This model would have little, if any, benefit in terms of building public sector skill base.

This model could negatively impact contributions to wider policy goals. The lack of public involvement means there is no way (other than through regulation) to control or influence provision of the service or to control pricing. Absent wider regulatory controls (e.g., through consenting or planning) there are no levers to ensure entirely privately-led schemes deliver against just transition objectives.

As these kinds of projects are scaled to meet the needs of the planned development which is generally controlled by the land owner / developer (and often required to install or connect to a heat network for planning reasons), they are less sensitive to demand risk. They are, however, unlikely to come forward without significant demand assurance stemming from the project itself.

This model supports the development of market-driven solutions. The development of private schemes will facilitate investment, thus supporting the transition to a self-sustaining market, particularly where schemes are less reliant on grant funding (for example, in areas where gas infill is not an option). The greater the number of successful heat network projects, even smaller private ones, the greater the level of market confidence. However, because schemes are generally limited to new development sites, private ESCo projects are likely to be limited in scale and to specific areas, and as such are unlikely to be sufficient to develop wider supply chains.

For the same reason, private ESCOs score badly for replicability across different geographies. Schemes are likely to be individual projects rather than large chain or groupings, and likely to work better on smaller scale, new developments; larger-scale projects incorporating existing public-sector anchor loads are more likely to necessitate some kind of public procurement. Private ESCo projects tend to happen primarily where the provision of (or connection to) a network is a requirement of planning policy. Such blanket, heat-network specific planning policies are not (yet) a feature of local area planning in Scotland.

This model also scores badly for interconnection and expansion. Delivery on a project-by-project basis, generally related to site ownership, does not encourage expansion. Generally developers are providing the network to satisfy planning requirements, not with a view to expansion. The heat network operator may or may not have the appetite or capacity to seek out future expansion or interconnection opportunities.

For similar reasons, this model is not anticipated significantly to encourage investment ahead of need. Projects would be expected to be scaled and designed to meet the committed demand; neither property developers nor heat network operators generally speculate on expansion.

These projects (as defined) would be classified to the private ESCO’s balance sheet.

7.4. Evaluation of DM4: Local authority led joint venture

In this model, a local authority procures a partner and forms a joint venture (JV) to serve an initial project (including one or more local authority anchor loads) and potentially additional projects and/or other energy projects within the local authority area. A full description of the characteristics of this model is set out in section 6.

Joint ventures (JVs) are legally more complex to deploy than a concession or Design & Build contracts, because corporate governance arrangements for the JV need to be agreed alongside commercial contracts for the initial project to be delivered. However, there are a growing number of established examples across the UK of JVs being used to deliver heat networks, including Midlothian Energy (a Joint Venture between Midlothian Council and Vattenfall Heat UK. JV models rely on existing, well understood corporate structures, and are no less compatible with forthcoming regulation than other models: consideration will need to be given to how any JV interacts with zoning and exclusive permitting. This model therefore scores positively for ease of deployment.

This model also provides good potential for private investment. Investors are generally attracted to the sharing of risk with the public sector partner that is inherent in a JV structure, when compared with, e.g., a service concession (although some private sector stakeholders were less interested in investing alongside the public sector than others). As heat networks are still considered by most to be higher risk projects, sharing risk (and returns) through a JV may offer up additional opportunities for investment that would not otherwise be attractive to investors. In addition, because a JV allows a single procurement to open up multiple, future projects, it provides a structure that will support future investment more easily. How easily that additional capital can be deployed will still depend on the individual arrangements: sometimes the need for both partners to approve investments can slow things down. A JV structure also allows a public sector partner to divest of their shareholding in the future should it desire, creating an additional opportunity for private investment (assuming the JV SPV was an attractive investment at that point).

A JV model can go a long way to simplifying delivery. Although JVs can take significant time to procure and there is no standardised approach to doing so, the process is robust and comprehensive and, after the procurement, there is the potential to deliver projects at scale and speed. The model allows both private and public sector to play to their strengths, working together to deliver both the initial project (which generally needs to be well defined in order to attract investors) and then to identify and develop subsequent projects. The procurement efficiency of a JV, which allows the JV to take forward multiple projects within a given area following a single public procurement exercise, is highly valued. If additional public sector stakeholders are also included in the JV structure or named on the Contract Notice, this can help to facilitate future connections / phases. If the procurement is properly structured, with transparency as to which public sector stakeholders will be able to connect their assets and take a heat supply from the JV, those stakeholders should not then need to run a separate public procurement exercise in order to do so.

The JV model also allows skills and capacity to develop in both the public and private sectors, striking a good balance between the respective skills and capacities of each partner. If the public sector partner can help to facilitate a pipeline of projects (not requiring separate procurements), the JV partner can provide staffing and resource to develop projects, and a supply chain to deliver them. Whilst the single local authority/private partner JV structure avoids multiple procurements for projects within a given geographical area, the JV’s activity and supply chain is still limited to that area, which will typically be a local authority administrative boundary.

The ongoing presence of the local authority, beyond the signing of an initial contract, may help manage and apply any policy-based outcomes or goals which have been agreed at the outset (noting that Board Member’s responsibilities are always to the JV company, not external entities) in addition to using the competitive procurement to mandate or compete minimum service levels or pricing policies. This combination should help to contribute to wider policy goals. The potential for multiple / aggregated projects could also help to reduce the cost of finance for the private sector partner (as larger investments can absorb due diligence costs more effectively), which should be reflected in customer charges. In addition, because JVs would be expected to unlock a larger number of projects/larger-scale projects, any positive contribution has the potential to impact a wider area. The downside of the JV model, and any model with a private sector partner, is that corporate interests and the need to make a return on investment must always also be prioritised, perhaps to a greater degree than if a public body was acting alone.

This model does not reduce demand risk more widely, but the opportunity to name other public sector bodies as buyers of heat in the Contract Notice may help to mitigate procurement-related demand risk for future connections / projects that the JV identifies, potentially making them easier to deliver.

The opportunity to build out multiple projects within an area from a single procurement helps take a longer-term and more strategic approach to developments. This may give rise to opportunities for the JV to aggregate projects, to benefit from portfolio risk management, and potentially reduce cost of finance, all of which support the transition to self-sustaining market (although the JV model does not in itself change the economics of individual projects). Although there is potential to ‘bundle’ projects for investment via the JV, this will still be limited by the geographic boundary of the procuring local authority, which will limit the opportunity for scale and hence cost reduction (except perhaps within the cities).

This model is replicable across various geographies, although if every local authority wished to pursue this model individually, it would generate a significant number of resource-intensive procurements, which may not all be able to attract adequate competitive tension if launched over a similar time period (bidders will focus resources on the areas with the most potential). This could be mitigated if some local authorities decided to procure jointly, which may be possible in a multi-authority JV or in DM10: RESCo model (see below). Once in place, the JV allows delivery of different solutions for different projects under the JV, making it more flexible than a concession model. The potential to aggregate projects or investments may create opportunities for smaller projects or extensions to existing networks that would not be commercial on their own.

The JV model is likely to support future interconnection and expansion. The JV private sector partner has a commercial interest in expanding schemes because it is their core business (unlike that of the public sector partner), meaning the JV should have the resources required to negotiate new connections. The public sector partner can then work to help facilitate delivery of those opportunities, through providing anchor loads or through engagement and influence with other local stakeholders (public or private). Inclusion of other public sector bodies in the Contract Notice would also support future expansions.

If the JV covered a large enough area over an extended period of time, this provides an opportunity to incentivise designs which support investment ahead of need. However, the JV model does not in itself achieve this. Whilst there is an opportunity to share investment risk and incentivise this kind of investment if the public sector partner is prepared to take a long-term view, both parties would need to approve individual project investments, which will always need to be commercial for the private sector partner. In practice, investment ahead of demand is usually achieved via subsidy (capital grants for future-proofing of energy centres / heat mains). This was the case for the Shawfair development, as the first project for (what became) Midlothian Energy.

Whether or not the SPV is classified to local authority balance sheets will depend on the individual structuring of the JV Co, including the parties’ respective shareholdings, rights, degree of control and approach to risk sharing. As the public sector partner is a local authority this is unlikely to score against Scottish Government capital budgets (other than any grants and potentially other financial support provided).

7.5. Evaluation of DM5: Community led project

In this model, a community leads heat network development and owns the network. The community then subcontracts the operation and maintenance of the network. These kind of schemes generally supply buildings within a community. A full description of the characteristics of this model is set out in section 6.

Although this model uses existing corporate structures, it scores neutrally for ease of deployment because, in practice, arranging appropriate governance can be complex to coordinate for a community group. It often requires the creation of a new community company, which brings additional legal complexities in relation to governance and the provision of guarantees, etc.

Although this model provides potential investment opportunities for community-sourced funding and finance, or companies interested in smaller scale investments, their potential for private investment is limited because it is likely to be an unattractive model for large-scale private investment, due to challenges with scale, control and transfer.

This model scores negatively for simplifying delivery. Although community companies do not have to run public procurements, and the community element is likely to be well received by local stakeholders, and may encourage engagement and cooperation from residential owners, these advantages do not outweigh the significant delivery and coordination challenges that small community projects often face. Projects are generally limited in size to networks serving a single, small community, which often has limited capacity to deal with interfaces and negotiations. Delivering community heat networks requires significant community organisation and commitment, which experience suggests is very hard to deliver in practice.

Although this model would develop skills and capacity at a community level, and may be able to support very local supply chains (local suppliers can be prioritised), these kinds of networks are unlikely to expand significantly or transfer local skills to other projects. This limits any wider benefits. These projects may also absorb scarce skills and capacity that would be better used elsewhere. For example, these projects are likely to represent an inefficient use of Government support time and resource (e.g. HNSU), because the amount of support they are likely to need would be disproportionate to the benefits delivered. We recognise however, as outlined below, that there may wider policy reasons for providing such support.

This model scores positively in respect of its potential to contribute to wider policy goals, because community led schemes are likely to be highly driven and focussed on wider policy objectives related to community wellbeing, fuel poverty and consumer protection. However, the lack of scale would limit any impact the project might have in any given area, and would also make it harder to support large scale training opportunities and supply chain development.

Small, community-led schemes may be slightly better able to manage demand risk among residential customers than other models. Community led initiatives are likely to be able to garner greater local awareness, trust and support for the heat network within the community and are therefore more likely to gain commitments to connect from households. However, any change in behaviours or attitudes is unlikely to be significant enough to reduce anxiety around demand risk to a degree which would impact the cost of finance. Schemes may end up being further reduced in scale as a result.

The nature of community led projects means they are less likely to attract significant private investment on a large scale, and are unlikely to give significant market confidence beyond a local level. As a result, they would make a minimal contribution towards a t ransition to a self-sustaining market, although they may perform well in practical terms at a local level, in giving communities confidence and engaging local suppliers.

This model is not replicable across geographies, because it will only work on small, local projects with a very engaged local community.

Although future interconnection and expansion would be driven primarily by the best outcome for consumers rather than a pure profit motive (which may remove some barriers to pricing/charging inefficiencies), in practice the integration of multiple community owned networks, or expansion of an existing one, is likely to be challenging in terms of the community’s capacity to negotiate and ability to invest and grow networks. Any expansions would most likely be very limited in scale.

Although community led planning may be open-minded and positive about future investment and open to investment ahead of need, the ability of any community led project to deliver this to any significant degree would be limited by both the size of the initial project and more limited access to capital that community led schemes tend to suffer from.

Assuming there was no public sector involvement or ownership at all, then projects delivered under this model would not be classified to any public sector balance sheets.

7.6. Evaluation of DM6: Unbundled model

This model involves the separate ownership of generation, transmission / distribution and/or supply assets. An example is where heat generators contract directly with customers and pay a use-of-system charge to the owner of heat transmission / distribution infrastructure. This model is in fact a family of models, some of which could be delivered to varying degrees within existing or other ‘new’ models. For example, there are some small-scale examples of this model (e.g., privately owned heat sources feeding into networks), or as a variation of other models (e.g., centrally led development, or funding of transmission networks by Government). For the purpose of evaluation, we have focussed on the use case involving separate ownership of all components from the outset.

A full description of the characteristics of this model is set out in section 6.

This model scores neutrally for ease of deployment. In addition to numerous international examples (e.g., Denmark, Netherlands), this model has already been deployed on a small scale in some places in the UK (e.g., Stirling, Glenrothes), and relies purely on contractual arrangements to manage interfaces between different elements of a network. A full version of this model, where transmission and distribution are separated, is potentially sensitive to forthcoming regulation in relation to permitting, as it is not currently clear whether permitting would support a distinction between these two activities. Standardising the process of buying heat from a third party source onto a network could help to deploy this model for projects reliant on a privately-owned heat source (e.g., industrial waste heat sources) but, if separate ownership of different parts of the network became commonplace, then additional regulation to manage interfaces may be preferable.

This model could increase the potential for private investment. The unbundled network model is already well-known to investors familiar with UK utilities’ markets (e.g., electricity networks). Unbundling may support direct, lower cost investment into the more traditional/well understood elements of heat networks, by separating that component from parts less familiar to investors. For example, if supply and distribution were separated, then investors more used to investing in customer facing businesses may be attracted to the supply business, whereas infrastructure investors may be more focussed on the physical components. In theory, the model also allows risk to be split by offering separate investment opportunities for lower and higher risk components. The potential for all of these advantages should grow with time, once assets are established at a greater scale, and are better understood by investors.

However, starting out larger investment on an unbundled basis is likely to be challenging, as the profitability of one element of the network is always likely to be dependent on another, e.g., a generation asset only has value if heat can be distributed to contracted customers, which would be beyond the generator’s control if the business was split. This interdependency risk may increase costs of investment, particularly at the outset before the supply business is established. The BEIS International Review of Heat Network Market Frameworks^[51] is not supportive of pursuing the unbundled model at this stage of the sector’s development. One investor stakeholder we spoke to stated that their long-term preference was only to own and operate heat network assets, and that they would prefer to transfer the supply elements (customer facing) to other entities who have more core business in this area, but accepted that this would not be possible until the market was much more mature.

The difficulty in pursuing this model from the outset (compared with the unbundling of mature networks later) means it scores badly in terms of its ability to simplify delivery. Complex contractual interfaces would need to be developed to manage risk across the various components. This model does not offer a clear path to large-scale delivery, unless (perhaps) Scottish Government was willing to fund and take the risk on certain transmission networks (with the option to divest in the future, once the network has been constructed and with stable revenue streams established). In some cases, for example where there is third-party waste heat available (e.g., an energy-from-waste facility or sewer source), a degree of unbundling is inevitable. The process of procuring such heat onto a network could be made easier if standardised.

This model would allow the development of some specialised skills in the different components, and could allow easier transfer of some existing skills and capacity from related industries (e.g., electricity supply companies could take on supply-side businesses).

This model scores neutrally in terms of its ability to contribute to wider policy goals. On the one hand, efficient private investment in particular types of assets could help drive down costs for consumers. If transmission networks were publicly owned, fuel-poor areas could be prioritised. On the other hand, interface risks could increase cost of capital, and the need to find profit for different businesses at each interface could increase costs for consumers. Any state involvement in a purely supply-side business seems less likely (as opposed to owning and operating whole networks, or transmission assets), therefore the ability to control pricing and protect consumers via controlling that element of the network is low.

This model is likely to be more sensitive to demand risk than other models, and so scores negatively. The risk is likely to be intensified because no single entity is able to manage all aspects of the operation: the capital-intensive elements (transmission) have no visibility of income from end users (generated by the supply-side business).

Developing transmission networks under separate ownership at the outset would require significant up-front investment and carry high demand risk as income streams would be dependent on activity by others. They would require significant public subsidy, which would not (at least in the short term) support the transition to a self-sustaining market. One investor/developer we spoke to was interested in ‘unbundling’ its networks in the future (by selling the supply elements of the business), but acknowledged that this would not be possible at the outset while the market is small and immature. There may be an opportunity to bring down consumer costs by leveraging institutional investment into certain assets in the longer term, by unbundling once networks of a sizeable scale have been established. However, this advantage would not be available at the outset because of the additional costs and unmanaged risks that arise from pursuing an unbundled model for new networks.

The BEIS study concludes that separating generation vs transmission/distribution vs supply would only be worthwhile on very large projects and geographies, as a result of the additional administrative costs associated with managing the interfaces and running the various components as separate businesses. The paper noted that studies from other countries also support this conclusion. As such, this model scores negatively for replicability, as it would only work for large projects or networks consolidated over a larger area (even once the riskier build phase has passed).

If one or more transmission networks were to be developed (likely by the public sector), this would support expansions and could provide opportunities for smaller distribution networks to connect into the transmission networks readily, reducing up front capital and risk to supplying new customers (although interfaces would need to be negotiated each time). This would support future expansions and interconnection, but would be reliant on a very significant up-front (at risk) investment in a transmission network.

Again, if large transmission networks were to be delivered, this would allow pipes to be significantly oversized to support delivery ahead of demand, albeit at cost and risk. Without some form of demand assurance, the inability to manage the downstream connection risk (either at point of construction or in the future) would increase the risk of this activity, making it more expensive.

The balance sheet treatment would depend on ownership of the individual assets. If some elements were publicly owned, this would be 'on balance sheet' for the relevant public sector organisation.

7.7. Evaluation of DM7: Merchant model

In this model a private sector heat network operator contracts with off-takers to supply existing buildings, without having either being appointed by a private sector landowner / developer in connection with a particular development site, or having followed a public procurement exercise. A full description of the characteristics of this model is set out in section 6.

In relation to ease of deployment, this model would score well if delivered in isolation, because it uses existing corporate structures, and the lack of public sector involvement in delivery role may streamline the process. However, when taken in the wider context of forthcoming regulation and a desire for coherent area-wide heat planning (via LHEES and city-wide heat network planning), a proliferation of this model may not be consistent with the regulatory and policy ‘direction of travel’. The future deployment of permits would be more challenging in areas where significant merchant networks have been established, because there would be an incumbent operator(s) in areas where local authorities would be seeking to confer exclusivity in order to benefit from a long-term, strategic approach. This will make permitting process more legally complex, as well as potentially undermining the commercial case for the permit (or the development of a strategic area-wide network by other means).

We consider this model is neutral in terms of potential for private investment. Privately owned structures are easy to transfer and exit from, and the lack of state involvement or public ownership would mean investments are not slowed by multiple interests. This model also offers opportunities to investors with a higher-risk appetite. On the other hand, it is hard to see how projects of this type can take place at scale (and offer larger investment opportunities) without any local authority involvement to commit anchor loads from the outset – public bodies may still have to procure the heat supply to such buildings. In addition, because a smaller number of investors are interested in projects with this kind of risk profile, these projects may only open up certain avenues of investment, with the potential to undermine others: merchant projects could undermine the commercial case for larger urban schemes by ‘cherry-picking’ key anchor loads, making larger schemes less likely. To create additional potential for investment, this kind of project would ideally be limited to areas where wider schemes are not being planned or prioritised at all, for example in smaller towns. There are also some micro-scale versions of this model, for example shared ground arrays, which could have a complementary role in less dense, suburban areas.

We have also scored this model neutrally in terms of its potential to simplify delivery. Although it does not rely on a public procurement to commence the project, legal complexities remain regarding procurement of heat supplies to public sector anchor-loads: it is unclear, in the absence of public procurement, what justification would be available to public sector bodies wishing to connect their buildings. There is therefore more potential for this model in areas where there is little or no reliance on public sector buildings. The lack of development of projects of this nature to date does suggest challenges around delivery. Challenges coordinating multiple stakeholders will still exist, and could be harder without any local authority involvement. Anecdotal evidence appears to suggest that getting anchor loads to commit remains challenging, even when construction has commenced, although this could improve over time. A proliferation of these projects may also have a complicating impact on wider delivery, by undermining city-wide permitting and procurement (see above under 'ease of deployment'). Such merchant projects are also likely to focus on targeting key anchor loads but without being incentivised (or required) to connect other, less commercially attractive buildings as part of the process, an outcome which can be avoided with a public-sector led procurement process.

This model will foster skills and capacity in the private sector, and would not put any pressure on limited public sector resources. However, a larger number of smaller merchant schemes (as opposed to larger, area-wide solutions) may be less efficient and could draw on private sector skills (e.g. advisory) in a less efficient manner.

This model scores poorly in terms of its ability to contribute to wider policy goals. The lack of public involvement (given no public sector sponsor or procurement) means there are limited ways to control or influence how policy objectives are met through delivery. UK-wide consumer protection for heat network customers are expected to promote fair pricing, and consenting and licensing may offer some scope for promoting wider policy objectives.

This model offers opportunities to investors with a higher-risk appetite who are less sensitive to demand risk, but fully private schemes are unlikely to come forward without significant demand assurance stemming from the project itself, which will limit their number and size. Given the limited risk appetite of most private sector investors, and absence of public sector involvement to mitigate risk, there is nothing inherent in the model which makes it less sensitive to demand risk.

Development of private schemes will support investment and provide confidence in heat networks more generally, if they can become viable on a stand-alone basis. This remains to be demonstrated. In any event, schemes are unlikely to be sufficiently numerous or of a significant scale to support a rapid transition to a self-sustaining market.

This model could be replicable across some geographies, but is likely to be limited to smaller scale projects, as delivering on a larger scale may necessitate some kind of public procurement. Projects are more likely to come forward individually, based on a specific geography, rather than in groups.

This delivery model itself assumes a degree of future expansion in order to make it profitable, although the lack of consolidated or consistent ownership of networks may make future interconnection and expansion more difficult on any significant scale. Any expansion is likely to be at risk of 'cherry picking' only the most profitable routes, rather than seeking to meet any wider needs.

This model (as we understand it) is more open than other models to accepting greater risk in advance of contracted demand from customers. This could, in principle, support investment ahead of need such as oversizing of pipes. However, without firm commitments from prospective customers, or firm regulation requiring or incentivising building owners to connect, it may be difficult to attract significant and sustained investment (particularly from third party investors).

By definition, these kinds of projects would be off any public sector balance sheet.

7.8. Evaluation of DM8: Centrally Led

In this model, Scottish Government (or an executive agency of the Scottish Government, or some other centrally controlled public body) would take the lead on development and delivery of projects, without need for local authorities to lead development. The central body could have initial ownership / part-ownership of projects (potentially alongside private sector partner(s)) but with potential for the onward sale or transfer of government stakes in projects once they are operational with established revenue streams. As with other delivery models, this model could be adjusted to perform slightly differently against a number of attributes. However, for the purposes of this evaluation, we have assumed that the central body takes full control of project development and delivery in particular areas, without the relevant local authorities being actively involved in the projects.

A full description of the characteristics of this model (as assumed for the evaluation) is set out in section 6.

In terms of ease of deployment of the model, it is likely that a separate entity (wholly owned by the Scottish Government) would be needed to ringfence projects / assets and allow for their onward sale. While a basic corporate structure could be set up, there may be a risk of ‘self-regulation’ in respect of the Heat Networks (Scotland) Act 2021, where Scottish Government is developing and operating networks, whilst holding regulatory functions in respect of heat networks. This issue could be reduced if a separate statutory entity was set up to hold the heat network assets and apply for the relevant approvals, but setting up an independent entity in this manner could require further legislation. A Scottish Government Business Case would need to be developed to take forward this model, which would take time, regardless of whether a separate statutory entity was preferred. Scottish Government already has experience of setting up arms-length infrastructure companies, for example the SFTi company, which was established by SFT to invest in community infrastructure projects under the ‘Hub’ initiative (the model on which DM10 (Regional ESCO) is based).

This model could increase potential for private investment, provided it successfully increased the number of projects being brought forward and the pace at which they were being delivered, when compared with a scenario where Government did not intervene. Through stakeholder engagement we had some strong, and largely consistent, feedback that greater direct Government involvement in schemes (although not necessarily via this model) would project confidence that heat networks are a viable, long-term solution which Government backed. This in turn may help reduce the perception of risk to local stakeholders and investors, and provide greater confidence, thus facilitating delivery. However, some investors (in particular those less active in or familiar with the Scottish market) were more negative about state involvement in delivery at all, with the view that ‘There are some things that Government is good at, and delivery is not one of them’ and ‘Government should regulate and set clear rules of the game, that’s all we need.’

Whether or not centrally led delivery would simplify delivery and increase the total number of projects coming forward is not definitive, and is likely to depend on how central resource was deployed, and any joint working arrangements with local authorities. For the purpose of the evaluation, we have assumed that local authorities would not be involved in project development and delivery in certain areas. A central body would be able to progress projects in areas where development is not being brought forward by local authorities at all, where local authorities do not have the capacity or the willingness to develop heat networks. It would also allow significant knowledge sharing and efficient allocation of resources, with a central body of expertise building up over time. We do have concerns, however, that central control over development could have a ‘shadow effect’, in which uncertainty is created about who is responsible for bringing forward projects, potentially undermining development of heat network opportunities that are currently attractive to local authorities. Local authorities may stop delivering and developing projects altogether, focussing resources elsewhere. A centrally led approach would be at odds with the current message that local authorities should be progressing heat network opportunities as one of the key outputs of Local Heat and Energy Efficiency Strategies (LHEES) and associated Delivery Plans.

Any central body would need to make difficult decisions about which projects to prioritise, and therefore which projects or areas to give less priority to. This would require considerable stakeholder and communication management and would inevitably be resource intensive, with the potential of distracting from delivery. In practical terms, we also see little benefit in divorcing the development and delivery of heat networks from local authorities, even in circumstances where the authority may not be behaving proactively. Local authorities are likely to be able to provide most of the anchor loads for a scheme, and will have essential local knowledge (including from the LHEES process) and connections that will be essential to make schemes a success. Centralisation of delivery would be counterproductive if the central body did not have sufficient skills and capacity. Projects are resource intensive, hence adequate resourcing would be required to progress multiple large projects in parallel. Although there are significant delivery benefits to be gained from centralisation, we believe full centralisation could have a negative effect on delivery.

Although there would be some benefit in creating a centralised core of expertise, we note that it may be challenging for central Government to recruit and retain the skills and capacity required to take on this work, and building up that capacity may take some time. This risk could be managed by providing significant, ongoing revenue resource alongside capital investment, and offering competitive salaries and opportunities which matched those available in the private sector.

Of all the models, centrally led delivery is likely to have the strongest link between central policy goals and delivery, meaning it has significant potential to contribute to wider policy goals. The degree to which this was deliverable in practice might depend on the models chosen by the central body to deliver individual projects: significant amounts of state control over projects would allow policy goals to be prioritised, whereas if private sector models such as concessions or joint ventures were adopted, policy drivers may become diluted by corporate priorities. We note that if a new entity was created (separate from Government), there is potential for the statutory responsibilities of the new entity and central Government’s policy priorities to diverge over time (e.g., keep consumer prices low instead of significant expansion).

Like most models evaluated, this option does not reduce (nor is it significantly less sensitive to) demand risk. It would give Scottish Government (or new entity) the opportunity to decide, at a project development stage, how much demand risk to retain. Retaining risk within the public sector may make projects cheaper and more deliverable from a private sector perspective. Local authorities may still need to be involved in committing anchor loads (though not in procuring heat networks), thus mitigating demand risk on individual projects.

If a centrally led body was able to unlock projects in otherwise stagnant areas, then it would begin to support a transition to a self-sustaining market. As noted above, stakeholders felt that greater Government involvement in delivery could send a strong supportive message to investors and the supply chain, but this would be sensitive to rapid delivery and demonstrating early success. Projects could be structured to provide an exit mechanism for any Government stake once projects are operational with established revenue streams, thus supporting future investment. Project and site selection would be key to avoid crowding-out private investment: central delivery would ideally be limited to areas where heat networks are deemed an appropriate solution, but development is not happening at pace and scale. This could mean a reduced likelihood of ‘quick wins’.

The kinds of projects that would be delivered under this model would depend entirely on the priorities and focus of the body (and deciding this could be very politically sensitive), but it scores well for being replicable across various geographies because the model could work for projects of different sizes, including those not otherwise commercially viable.

Looking to the future, significant centrally led involvement could offer the strategic advantage of identifying opportunities for interconnection and expansion. If the central body owned or held a stake in most of the projects it delivered (which may not necessarily be the case), then that control could be used to encourage expansion / interconnection to take place where it was strategically beneficial to do so, rather than being purely profit driven. If the central body took on significant ownership and was willing to take on greater risk, it could also encourage and fund investment ahead of need. However, both retaining a stake in projects (in order to exercise greater influence and control) and investment ahead of need would require significant ongoing resource and capital budget. Projects with a significant Government stake (or other form of control) would be more likely to be classified to Scottish Government’s balance sheet, and therefore reduce funds available to be spent on other priorities.

7.9. Evaluation of DM9: Local Authority led projects; Scottish Government stake

In this model, local authorities lead the development and delivery of projects, with Scottish Government / central support and co-investment (which may be in addition to an element of grant), in a joint venture arrangement. Scottish Government would have part-ownership of schemes, but with potential for sale / transfer of government stake once scheme is established. JVs could be bipartite – local authority & Scottish Government, or tripartite - local authority, Scottish Government and private sector. Please see Figure 3 (section 8) for an illustration of this. A full description of the characteristics of this model is set out in section 6.

There are some challenges in relation to ease of deployment for this model. While a basic corporate structure could be set up, with Government holding investments directly in the joint venture, there may be a risk of ‘self-regulation’ under the Heat Networks (Scotland) Act 2021, where Scottish Government takes a stake in projects, whilst also holding regulatory functions in respect of heat networks. This would be more likely for projects where the Scottish Government held a majority stake. This issue could be reduced if a separate statutory entity was set up to hold the heat network assets and apply for the relevant approvals, but setting up an independent entity in this manner could require further legislation. A Scottish Government Business Case would need to be developed to take forward this model, which would take time, regardless of whether a separate statutory entity was preferred. Scottish Government already has experience of setting up arms-length infrastructure companies, for example the SFTi company, which was established by SFT to invest in community infrastructure projects under the ‘Hub’ initiative (the model on which DM10 (Regional ESCO) is base).

This model could provide additional potential for private investment, particularly where a private sector partner is also involved via a tripartite JV. Similarly to a local authority/private sector JV, the sharing of risk inherent in the structure can be attractive to the private sector, helping to de-risk investments compared with, e.g., a service concession. This advantage may be further increased if one of the other parties in the JV was Scottish Government. Involvement of Scottish Government may give local authorities the confidence to bring forward schemes that would otherwise not happen, creating additional investment opportunities. Through stakeholder engagement we had some strong, and largely consistent, feedback that greater direct Scottish Government involvement in schemes would project confidence that heat networks are a viable, long term solution which Scottish Government backed. This in turn may help reduce the perception of risk to stakeholders and investors, and provide greater confidence, thus facilitating delivery. Scottish Government investments could be structured in a way that does not disincentivise other investment (e.g., making Scottish Government investment subordinate to that of other investors), although determining what an optimal structure would be requires further detailed consideration. A JV structure would also allow Scottish Government to divest its shareholding in the future should it desire, creating an additional opportunity for private investment (assuming the JV SPV was an attractive investment at that point).

The downsides of this model, in relation to the potential for private investment, are that governance structures for JVs can be complex, with investments generally needing to be approved by both the JV and separately by its public and private shareholders. This complexity would be multiplied by having Scottish Government as a third shareholder, with potential for approvals needed from three separate organisations. A JV would therefore require sufficient delegated authority from its shareholders to allow efficient decision making. Ensuring Scottish Government could hold and manage multiple shareholdings efficiently would be key to success of this model. As noted above under DM8, some investors (in particular those less active in or familiar with the Scottish market) were more negative than others about state involvement in delivery, and less interested in sharing risk/return with government. A minority we spoke to held the view that, ‘There are some things that Government is good at, and delivery is not one of them’ and ‘Government should regulate and set clear rules of the game, that’s all we need.’ There is also a risk that the market is seen as being "tied up" by Scottish Government stakes and therefore risks crowding out other investment. Careful selection of which projects to invest in may help overcome this.

If the joint venture was bipartite, that is between the local authority and Scottish Government only, then the potential for private investment would score the same as DM1, as it would be a fully publicly owned scheme.

In terms of simplifying delivery, a tripartite JV involving a Scottish Government stake has the same advantages and disadvantages as DM4 (Local authority led joint venture). In addition, the involvement of Scottish Government could help to facilitate delivery, by providing a more direct connection between Scottish Government’s heat network policy priorities and delivery. It may also make it easier for centrally-owned anchor loads to be involved in projects. The additional resource and expertise offered by a Scottish Government shareholding could also add value over time, as experiences and best practice from other projects across the country could be shared via a single central conduit. Having Scottish Government involved in tripartite JVs may also have some delivery disadvantages: procurement may become lengthier and more complex by involving Scottish Government, unless a separate entity (or internal unit with significant autonomy) could be created. Similar ‘drag’ from central involvement could also carry into delivery, depending on the governance arrangements – see above under ‘potential for private investment’. Work would be required to ensure that those managing central shareholdings had the resource and capacity needed to add value.

In terms of a bipartite, public-public JV, the delivery advantages of bringing in the private sector (as described at DM4) would be lost. However, there may still be delivery benefits in providing ongoing, central support to local authorities that want to own and operate schemes (or where schemes are not attractive to the private sector), but where the local authority (or Scottish Government, who would otherwise be providing grant only) are not satisfied that the local authority has the expertise to take the project forward. Ensuring that the support added value would be key: one stakeholder we spoke to regarded central co-investment as helpful, but only if the practical support associated with that investment provided continuity and additional commercial or technical skills, stating “We don’t need civil servants who don’t understand heat networks and who change every 6 months”.

In terms of skills and capacity, a tripartite JV with a Scottish Government stake under this model has the same advantages and disadvantages as DM4 (Local authority led joint venture). In addition, an ongoing role for Scottish Government in multiple projects would help to develop skills, knowledge and best practice centrally, which can then be shared with other projects.

The ability of this model to contribute to wider policy goals would be similar to DM4 if the JV was tripartite (i.e., with the private sector). Having an additional public sector ‘voice’ involved in corporate decision making may further support policy priorities where there is appropriate discretion within the corporate framework (as explained in DM4, above), although it is possible that the two public sector voices may not agree on how best to apply those priorities. If the JV was bipartite (without the private sector), then the ability to contribute to wider policy goals would be more similar to DM1.

The model does not itself reduce demand risk. However, involving local authorities in the delivery of schemes (as opposed to proceeding without them as under DM8) can help to manage demand risk, as local authorities can commit anchor loads. In a JV model, this commitment can be ongoing and may help to deliver subsequent schemes. A bipartite, public-public JV may be able to take more demand risk, because risk is shared between the local authority and Scottish Government, and is not reliant on private sector investors. In practice, this will depend on the risk appetite of the investing authorities.

A Scottish Government stake in projects may help the transition to self-sustaining market, if it provided confidence of project pipeline to supply chains, as they observe greater Scottish Government commitment to heat networks. It may generate additional projects that would not come forward at pace if local authorities are asked to proceed alone, creating additional investment opportunities. Providing equity and loans, as opposed to grants only, also allows Government to ‘recycle’ funding into future projects, which may support additionality over the long term. The success of this intervention, and the degree of confidence it provides to the market, will be entirely dependent on early successful delivery and positive case studies.

This model scores well for being replicable across geographies. A consistent central role in multiple JVs across Scotland helps make them easier to replicate. Most of the advantages of this model are similar to those of DM4, including that once procured, the JV allows delivery of different solutions for different projects under the JV, creating opportunities for smaller projects or extensions to existing networks that would not be commercial to bring forward on their own.

This model also scores well for supporting interconnection and expansion. Assuming a tripartite JV involving the private sector, the advantages are similar to DM4, but the additional involvement of a central body could provide a degree of strategic oversight, and an ability to promote interconnections. It may also help support expansions and interconnections which are not necessarily commercially attractive, but which would have wider societal or policy-based advantages, albeit this influence would always be tempered by the need to make decisions in the interests of the JV company, and by the commercial interests of the private co-investor. Involving local authorities in the JV structure, as opposed to proceeding without them as in DM8, will also ensure that local objectives can be reflected in any expansion proposals, ensuring better alignment with LHEES. In a bipartite JV (not involving a private co-investor), performance against this attribute would be poorer, because experience suggests that private sector involvement is often key to treating networks like businesses and driving expansion opportunities.

A Scottish Government stake in projects may help to support investment ahead of need, provided such investment was also in the interests of the JV as a corporate entity (Scottish Government’s primary duty in the JV governance would be to the company). However, while Scottish Government may be prepared to take a long-term view on demand, all parties need to approve individual project investments, and the associated risks may be unattractive to the local authority or any private investor. In practice, such investment would likely need to be supported by subsidy or by Scottish Government carrying most of the risk.

There is potential for this model to be off balance-sheet if Scottish Government ownership and control was limited. However many of the benefits of model come from (or are significantly enhanced by) Scottish Government having control or heavy influence over decision-making within the JV.

7.10. Evaluation of DM10: Regional ESCo (“RESCo”)

In this model, local authorities and other public bodies (e.g., NHS, universities / colleges) come together on a regional basis and jointly procure a private sector delivery partner for each region (similar to the Hub model, which was established in Scotland to deliver community infrastructure). The local public partners then use the delivery partner to scope, design and deliver projects, according to pre-defined contracting structures, drawing on the delivery partner’s supply chain. A full description of the characteristics of this model is set out in section 6.

This model has been given a neutral score for ease of deployment. A similar model has been successfully deployed for Hub, and learning from the Hub initiative could be applied to the development of a RESCo model. This model would not require new legislation, although if Scottish Government wanted to hold a stake in the RESCos (like SFT does in Hub via SFTi) then similar consideration would need to be given about how that would best be done, for the reasons described in DM8 and DM9. It would however be a new model for heat networks, and would require a number of regional procurements (for which a single suite of documents could be developed). Procurement would be preceded by a significant development period, including more detailed market testing of the concept. Hub took around two years to establish, although some efficiencies may be possible in delivering this model if learning from Hub can be applied. Consideration would need to given as to how this model could best align with the regulatory regime, in particular permitting (as each RESCo would need to offer some initial exclusivity to the private sector partner, in order to make the proposal commercially attractive).

This model provides good potential for private investment. The regional scale of the RESCo will provide multiple potential projects for the private sector partner to develop, using finance from its own balance sheet and third-party investors (where the public sector partners did not wish to provide funding for projects themselves). As well as being able to access finance, the RESCo partner will be incentivised to identify investment opportunities, bringing forward projects more readily than if the public sector had to generate these projects alone. The scale and duration of any exclusivity offered to the RESCo partner would influence the potential for investment at the outset.

This model scores very well against the simplifying delivery attribute for a number of reasons. Firstly, like a local authority led JV, a single procurement for the RESCo partner would unlock multiple projects, and would do so over a larger area, because the procurement process would involve multiple public sector bodies within a region. The ‘partnering’ nature of the model, where other public sector bodies have a stake in the RESCo (in addition to being named on the Contract Notice) delivers procurement efficiency but should also help to catalyse development, due to a larger number of public sector entities (with responsibility for many potential anchor loads) in a more proactive role. The RESCo could act as a ‘one stop shop’ for participating public bodies in the region to be able to purchase from, making it easier to develop projects using the private partner’s management / design services. The heat network delivery partner (which may be the main RESCo partner, or within the RESCo partner’s supply chain) would also be able to seek out wider development opportunities, including larger networks which connect buildings beyond those anchor loads offered by the public sector partners. The model could accommodate multiple contracting structures, allowing choices to be made on a project-by-project process. Selecting which structure to use may initially create complexity, but it is anticipated that best practice/preferred approaches would emerge over time. We know that the existing Hub model has significantly improved delivery of community infrastructure, but more detailed review is still required to determine whether the same degree of benefit would be achieved in the context of heat networks.

This model also scores very well against the skills and capacity attribute. Like all the JV models with a private partner, it strikes a good balance between the respective skills / capacities of the public and private sector partners, as described in DM4. The advantages are similar to those for DM4, with three main additional advantages: firstly, the regional nature of the proposal would allow further efficiencies in deployment of private and public sector skills, and provide easier access to heat network expertise (including to expertise accessible at an early stage in project development) for public bodies other than just local authorities. Secondly, each RESCo could have its own team with priorities focussed on the local area, developing expertise that can be deployed across the region. Thirdly, the involvement of more public sector bodies within the RESCo creates an opportunity to build local relationships and create platforms for sharing learning. The benefit of this has been very apparent in Hub, which in some regions are actively used to bring partners together on a range of topics. The regional nature of the RESCo would also have benefits for the wider supply chain, by providing a pipeline of work that does not require a public procurement to access in each case (as the supply chain is managed by the RESCo partner). There is evidence that the Hub model has had a very positive effect on supply chains.

The advantages and disadvantages to this model in respect of contributing to wider policy goals are similar to those identified in DM4 and DM9, except that benefits could be spread over a wider area, and there would be a larger number of public sector stakeholders involved. This could bring benefits, as greater awareness within the RESCo of how policy issues affect different public bodies may add value, but could also bring complications in setting priorities.

The proactive involvement (as distinct from the more passive step of being named on a Contract Notice) of multiple public sector bodies in the establishment of a RESCo could represent early, “in principle commitment” of anchor loads from all of the bodies. This more coordinated approach to the delivery of heat networks across a wider area to serve all participating public bodies makes this model slightly better able to cope with demand risk than most others.

This model would support the transition to self-sustaining market. The creation of a buying framework would make it easier for public authorities to develop heat networks (or facilitate the development of networks for which they provide anchor loads), generating more projects and investment opportunities more easily. The ‘partnering’ element of the model encourages the public sector partners to invest time and capacity into considering the development of heat networks in their area, which is more likely to generate projects: a number of the Hubs have fostered successful collaborative working to develop more efficient infrastructure solutions, and a RESCo could potentially deliver similar results. The regional nature of the partnership may give rise to opportunities for the RESCo to aggregate projects more easily, to benefit from portfolio risk management, and potentially to reduce the cost of finance. A larger number of public sector stakeholders making area-specific commitments to deliver (or connect their assets to) heat networks should build confidence for supply chains, which may be more encouraged to invest in their own businesses as a result. Specification of areas to be targeted and accompanying timelines could further enhance this benefit.

This model scores well for being replicable across multiple geographies. RESCos could deliver across all geographies and sizes of project (where heat networks are an appropriate technical solution), supporting smaller projects that may not be deliverable as stand-alone projects. This is because procurement, supply chain, governance and access to finance via the private partner would already be established, so the cost of developing and delivering new proposals is relatively low. Projects which would ordinarily be unattractive due to prohibitive bid costs could become more viable. It also offers a simple delivery route for projects which are not investible for the private sector, but which a public sector partner wishes to fund themselves. The model would also allow greater standardisation of approach across the region, providing access to the same skill set and supply chain for all projects.

This model could also have a positive impact on interconnection and expansion. Like all of the models involving a private sector partner, we would expect to see the RESCo partner actively looking for opportunities to expand and interconnect schemes in their area. Regional perspective and the potential to coordinate multiple, smaller schemes may make it easier than a local-only JV to identify opportunities for expansion. The commitment of other public sector bodies in the RESCo can also provide practical support for future expansions, because expanding to additional anchor loads may not require additional procurements.

The RESCo model scores neutrally in terms of investment ahead of need. The model does not in itself facilitate installation ahead of demand. Like all the JV models, whilst the public sector partners may be prepared to take a long-term view on demand, the RESCo as a whole (including majority private sector voting rights) needs to approve individual project investments. However, higher levels of cross-public sector involvement may allow for greater reliance on longer-term connection opportunities without signed contractual arrangements. For example, knowledge that hospital boiler plant will need to be replaced with a zero emissions system by a certain date, together with the relevant NHS Board involved in the RESCo, should help to de-risk investment ahead of need in associated infrastructure.

Whether or not the RESCos would be classified to central Government balance sheets will depend on the amount of shareholding and the degree of control. A significant amount of work and preparation went into the Hub balance sheet treatment, the lessons from which could be applied here.

7.11. Evaluation of DM11: Public Private Partnership (PPP)

Under a PPP, the heat network is operated by the private sector under a long-term contract tendered by a public body. The public sector sponsor retains the majority of demand risk, but availability risk lies with the PPP contractor (in contrast to a concession, where under our definition the private sector takes the demand risk). A full description of the characteristics of this model is set out in section 6.

We have scored PPPs neutrally in terms of ease of deployment. There is significant experience in deploying PPPs in Scotland for other sectors (we are not aware of any examples of heat networks), and the process is relatively robust, albeit lengthy and expensive for both public sector and bidders. There are some additional complexities when tendering for a PPP compared with a concession. In particular, a PPP involves significant technical and financial due diligence by lenders in addition to the procuring authority and private contractors, making the bidding process longer and more expensive. This model is not at odds with the Heat Networks (Scotland) Act 2021 but, like most models, consideration needs to be given to how commercial tendering processes can best align with/feed into Heat Network Zones and the subsequent issuing of permits.

In theory a PPP provides potential for private investment, but in practice it would be challenging and expensive to finance PPP heat networks where there is any degree of demand risk. This inflexibility means that a PPP would only have an acceptable risk profile for small heat networks with a limited number of off-takers and limited prospects (or intention) of expanding the network. However, the need to recover the significant bid costs associated with a PPP would make small-scale projects too expensive to finance.

Given it is hard to see PPP heat networks being possible at any kind of scale, we believe an attempt to encourage this model would not simplify delivery. They take longer to procure, require the involvement of more advisors and cost more. We see no appetite for this model for UK heat networks.

Due to the extended procurement timelines and additional input needed from advisors to deliver a PPP contact, it is likely to place an additional stress on an already thin advisory market, without delivering a greater number of projects (compared with, say, a concession). Once appointed, the PPP partner would run the project. Local authority resource would be required for contract monitoring. The PPP model offers no benefits in this regard over a standard concession. We have scored it neutrally for supporting skills and capacity.

This model is not likely to help contribute to wider policy goals. Although outcomes can be driven via the PPP contract, securing those outcomes over the long term via an initial contract negotiation is challenging. It is difficult to draft contract conditions in such a way that allows them to be sufficiently flexible and monitored over time, without such flexibility being perceived by bidders as a risk.

It is inherent in the PPP model (as defined for this paper) that the public sector would retain demand risk of the project. Unless the network is small (and therefore not cost effective for the PPP model), the public sector may not be well placed to manage that risk.

Greater use of this model is not likely to make a significant improvement towards the transition to self-sustaining market. It is predicated on third party investment, which is currently limited in heat networks in the UK today. The PPP model is likely to be a complex and slow way of attempting to encourage that, because it would take time to get banks comfortable with the nature of the asset to the degree required for a PPP, which in turn would slow down the transition. As PPPs are more expensive to procure, they are likely to result in a more expensive overall solution, as bid costs are inevitably passed on, making them more reliant on subsidy.

We do not consider this model to be replicable across multiple geographies or project sizes, because networks will likely have to be small and relatively self-contained in order for the risks to be manageable to an extent that allows them to be delivered via a PPP. As noted above, the expensive procurement process would not be justified for small networks.

Unlike a concession, there is limited incentive for the PPP contractor to seek out interconnection and expansion opportunities. PPP contracts are notoriously difficult to change, and are (by their nature) more suitable to stable assets which remain the same for the contract term. The PPP model therefore scores negatively against this attribute.

By their nature, PPPs are not well suited to projects which involve significant change over time. Hence demand would need to be secured ahead of contract signature. For this reason, this model is unlikely to support any investment ahead of need.

PPPs have the potential to be off the procuring authority’s balance sheet if sufficient ownership, control and risk lies with private sector.

7.12. Evaluation of DM12: Regulated Asset Base (RAB)

The Regulated Asset Base (RAB) is a private sector ownership model in which heat network assets are constructed, owned and operated by a monopoly supplier on a long-term basis. Investment plans, operating performance and returns (which are capped) are subject to regulatory oversight. The model is intended to incentivise private investment in large-scale heat networks, with a cost of capital comparable to other regulated utilities. A full description of the characteristics of this model is set out in section 6.

Although this model is well established in the UK in other sectors and has potential long term value, it would be a new model for heat networks. The regulatory system associated with the model is not contemplated by existing or forthcoming regulation, and hence would require a significant change in approach. Its application would probably be limited to cities. Deployment of a new RAB for heat networks would require legislation. Its introduction could require a mix of devolved and reserved powers (e.g., to regulate pricing), hence the agreement of the UK government would also likely be required. We have therefore scored this model negatively for ease of deployment, as it is not deliverable at this time.

This model would , by definition, involve significant private investment. It provides for customer charges to be set periodically in order to meet a capped return approved by the regulator. There is significant potential for private investment in heat networks in cities, which is where a RAB would likely be targeted. However, the minimum efficient scale for a heat networks RAB to be cost effective (taking into account the anticipated level of returns and regulatory costs) is not known. Appetite for a new RAB has not been formally tested with the heat network market or investors, hence is unknown. However, stakeholder feedback indicated a growing acceptance that, if heat networks are to achieve their long-term potential (circa 15-20% of heat supply by 2050), then a ’long term’ delivery model for heat networks would likely involve consolidation of asset ownership, underpinned by central regulation (which would be required, given the monopolistic nature of supply via heat networks).

The RAB model scores neutrally for its ability to simplify delivery. Once the structure was established, it would be easier to roll out projects and expansions in areas which the RAB covered when compared with taking forward individual projects However, a RAB would probably only be applicable in a very small number of geographies, primarily large cities, because a large customer base is required to spread investments. This means that other delivery models would still need to be used to deliver heat networks in other areas. It also means that there are unlikely to be enough heat network assets to support transfer to a RAB for some time (even if the RAB were to be ‘national’).

A RAB could result in significant investment, having a positive impact on the development of associated skills and capacity in the areas in which it was operating. Like other models with significant private sector involvement, the RAB operator would be able to bring resources and capacity. Even if the RAB remained in public ownership, it would be expected to operate as a business (similar to Scottish Water), and so would be resourced accordingly.

The RAB model has potential to contribute to wider policy goals. The intention would be that the RAB operator would have a low cost of capital, given that (regulated) returns can be achieved, which should be attractive to institutional investors (as evidenced by other RABs in operation). This low cost of capital should be reflected in customer charges. The role of an ongoing regulator should also help to support consumer protection, because the regulator would also scrutinise service delivery.

This model is the least sensitive to demand risk, because the customer base/stakeholders funding the RAB ultimately carry the demand risk for any investments. The model effectively socialises the risk by sharing it among an assumed large customer base, reducing risk for the RAB operator by allowing it to set and recover charges via customers/stakeholders, based on an allowed regulated return. However, the model can only manage demand risk effectively if there is a large customer base among whom to spread the risk (see simplifies delivery commentary, above).

The RAB model would support the transition to self-sustaining market. The opportunity to build out multiple projects within an area helps take a longer-term and more strategic approach to developments. This will create supply chain confidence and may give rise to opportunities for the RAB operator to reduce the cost of finance via larger-scale investments, which in turn would improve project economics and hence support the transition to a self-sustaining market. However, the RAB model does not in itself change the economics of individual projects, and investments (although potentially capable of being bundled) will still be limited by the geographic boundary of the RAB's operation, which will likely be limited to large cities. There is a potential spill-over effect from a RAB’s well-developed supply chains, which could increase the skills and resources available to projects outwith the RAB area(s).

The RAB model does not score well for replicability, because it requires a large customer base to support it. As such, unless a single, national RAB was developed, this model is likely to be of potential application only in a small number of large cities.

The RAB model scores very well for its ability to support the interconnection and expansion of projects. A RAB operator would have a commercial interest in developing new schemes, and expanding existing schemes within its area of operation. This is because the operator's return on capital employed is regulated, meaning it can only increase the quantum of returns by expanding its asset base. However, because the RAB model is likely to have potential application only in a small number of large cities, it would not support expansion and interconnection outside of these areas.

The model also scores relatively well in terms of investment ahead of need. A RAB operator is more likely to make these kinds of investments (provided they have been approved by the regulator as part of the periodic investment planning cycle), because customer charges can be set to achieve the allowed returns on investments,. However, regulators tend (over time) to be focussed on consumer protection and keeping consumer costs down, and can be reluctant to approve significant spend to deliver ahead of need, due to concerns about stranded assets. This is often symptomatic of how strategic objectives for RABs have been written, which reflect the need to balance multiple factors (e.g., the need decarbonise versus the cost to consumers).

If the RAB was privately owned, then it would not be classified to public sector balance sheets. If it was owned by Government, then it would have a significant impact on public sector balance sheets, given the scale of the assets under its control (similar to Scottish Water).