The role of planning in delivering low-carbon urban infrastructure

Introduction

Cities currently account for 80% of anthropogenic CO 2 emissions produced globally (Hornwegg et al, 2011). They consume around 60% of global energy, and are largely dependent on fossil fuel (IEA and OECD, 2008). This makes cities particularly vulnerable to hikes in fuel price and energy embargoes. It also means they are the single largest contributor to climate change.

Various technical solutions can be adopted to reduce fossil-fuel dependency and carbon emissions from cities including: low-energy and passive buildings, smart grids, trigeneration, district heating and cooling systems, wind farms, solar arrays, small-scale hydropower plants, biofuel, and closed-loop systems. Low-carbon infrastructure (LCI) can be integrated into city regions, at a range of scales assisted by planning policy and process.

The role of planning in delivering LCI is illustrated in this paper using international examples of best practice; these include: Freiburg (Germany); Malmö and Stockholm (Sweden); San Diego and Boulder (USA); and London (UK). All these cities have successfully integrated some LCI into their existing urban fabric. This has been assisted through the planning process.

Of course the planning systems (processes and regulatory controls) in each location are rather different, thus the transferring of planning approaches between contexts may be difficult. However, by studying a range of contexts and planning systems one can begin to see the range of approaches that might be adopted to enable integration of LCI in cities, the strengths and weaknesses of the approaches, and the contextual factors that enable their success.

These examples are drawn from the Zero Carbon Realities (1) research project. Detailed case study analyses were completed in the cities listed. In-depth interviews with infrastructure providers, energy experts, and urban planners were conducted to determine planning's role in the delivery of LCI in each city. A range of secondary data sources (2) were used to triangulate the findings.

This paper presents three different planning approaches (illustrated by European and American examples) for encouraging the integration of LCI into cities: the collaborative, systemic, and market-shaping approaches. The contexts in which each planning approach is most effective are established and strengths and weaknesses are identified. A discussion follows, on how planning could be used to provide a 'protected space' in which low-carbon systems can develop.

A role for planning

The literature has identified various roles for planning in enabling the delivery of the urban low-carbon agenda (Bulkeley and Kern, 2006;Campbell, 2008;Davoudi, 2009) and more specifically LCI (Crawford and French, 2008;Day et al, 2009). The only complete typology thus far developed has been attempted for the UK context (Davoudi, 2009). This simple typology outlines the role of planning in delivering key climate-change policies (table 1). Although the typology is useful, it is perhaps over simplified. Its two key weaknesses as a (1) The Zero Carbon Realities research project was conducted between 2007 and 2011. The project was funded by the Grand Challenge Prize. A full account of methods and results from the project can be found in Zero Carbon Homes: A Road Map (Williams, 2005).

(2) These included spatial, energy, resource, and transport plans; planning committee reports; transcripts from stake-holder meetings; local statistics; consultants and academic reports. typology for planning approaches delivering LCI are that it is UK-centric and that it focuses on the delivery of general climate-change policy rather than the specific delivery of LCI. Much of the discussion in the past two decades has revolved around the role of planning in delivering successful low-carbon transport systems. It has highlighted how spatial planning can influence urban form and regional development patterns in a way that reduces the need for travel, and promotes a modal shift and the adoption of appropriate LCI (Banister et al, 1997;Newman and Kenworthy, 1999;Schwanen et al, 2001;Stead, 2001;Williams, 2005;Williams and Banister, 1998).

Planning's role in supporting energy supply infrastructure has been less well documented. A great deal of the UK planning literature has focused on the barriers the planning system has created for the development of large-scale renewable energy projects (Foxon et al, 2005;Kellet, 2003;Painuly, 2001). More recently a discussion has developed surrounding the success in using planning as a tool for enabling decentralised forms of renewable energy (Allen et al, 2008;Crawford and French, 2008;Day et al, 2009;Williams, 2010). Others have discussed the value of using planning as a tool for delivering low-carbon buildings compared with mandatory and voluntary building codes (Lowe, 2007;Lowe and Oreszczyn, 2008;Williams, 2008).

The weakness of using planning as a tool to support the delivery of LCI has been partially explored in the UK context. Here the policy-implementation gap appears to be large. Planning authorities suggest that competing priorities and lack of resources for assessing nonstandard infrastructural systems are responsible for limited delivery (Allen et al, 2008;Campbell, 2006;Crawford and French, 2008;Day et al, 2009;Lowe, 2007;Lowe and Oreszczyn, 2008;Williams, 2008;.

In addition, cities are influenced by global processes, institutions, networks, and trends rather than by urban governance (Amin, 2007), which undermines the role of urban planning in the delivery of LCI. This situation has been further exacerbated by the deregulation and privatisation of network services (Healey, 2006;Lefèvre and d'Albergo, 2007;Pierre, 1999), which has dramatically weakened the power of municipalities to require the inclusion of LCI in cities.

Systematic urban planning and regulatory functions of urban government have partially given way to project-driven practices, entrepreneurial and collaborative approaches to governance (Harvey, 1989;Healey, 2006). Thus, it is likely that new planning approaches for the delivery of LCI have developed to reflect this change. Planning can create 'protected spaces' in which interactive learning processes and institutional adaptation can take place, helping new infrastructural systems to become competitive . This is a top-down process with the emphasis on technological experimentation rather than coevolution of sociotechnical systems. Alternatively, a more collaborative form of planning could provide support for niche innovation originating within the community [as suggested by Harvey (1989) and Healey (2006)], which would enable the coevolution of sociotechnical systems.

The local context is likely to influence the way in which new infrastructural systems develop (the processes and stakeholders involved). However, there is a tendency to look at technological innovation processes without reference to spatial context (Berkhout et al, 2004;Monstadt, 2009;Raven, 2007;Rip and Kemp, 1998;Smith et al, 2005). Local urban management approaches will also influence the type of systems and the processes through which they evolve. Thus, if the planning system is to be effective in the delivery of LCI it will need to reflect the local context in which it operates. These ideas in relation to LCI need to be developed.

A review of the literature suggests that plans, policy, and process could help to deliver LCI in three ways.

(1) The process itself could be used as a vehicle for consultation and for encouraging collaboration between stakeholders in the development process, which would enable the delivery of LCI.

(2) Strategic plans can be used to coordinate LCI, urban form, and new development.

(3) Regulatory requirements could be placed on new developments to include LCI.

However, most of the literature is UK-centric and is focused on delivering more-generic low carbon objectives, rather than the delivery of LCI. Thus a more thorough investigation is needed. The Zero Carbon Realities research project suggested that there are some subtle but important differences in the approaches needed to provide bespoke LCI systems in cities across a range of international contexts.

The typology

The research identified three planning approaches to LCI integration in cities-collaborative, systemic, and market shaping. In the following discussion this typology will be presented.

The collaborative approach

The collaborative approach has been used to achieve environmental objectives in cities where there has been community support (Innes et al, 1994;Randolph and Bauer, 1999;Selin and Chevez, 1995). It includes multiple stakeholders from the community in the development process. Community stakeholders are involved directly in dialogue with public agencies. They are engaged in defining the parameters and potential options for new development (including the green agenda).

In some instances they may also be involved in the financing, design, construction, and operation of the development (coprovision). Involving a diversity of stakeholders in this process is likely to create conflict. However, if it also produces constructive debate it can increase understanding amongst stakeholders and support for environmental projects.

The collaborative planning process helps to build capacity within the community (Healey, 1997). Participants learn about the technical, organisational, financial, and political aspects of development. Stakeholders cultivate their leadership skills and learn to prioritise their goals (including environmental goals) as a group. New partnerships form between stakeholders enabling the delivery of projects (particularly useful for community infrastructure). The community is empowered and takes ownership of the project, and thus feels responsible for its successful delivery and ongoing operation. In some instances the community becomes the coprovider of LCI and services.

The planning authority acts as the main catalyst in this process. It identifies key stakeholders and ensures that all are involved in discussion and decision-making processes. Planners facilitate this process by providing guidance, consultancy, and expert opinions. They also mediate and help to resolve conflicts between stakeholders, as well as setting clear objectives for development. However the community has the final authority to make decisions. This approach requires full commitment from a municipality for it to be successful (Innes and Booher, 2000).

The collaborative approach goes beyond the normal levels of public involvement in the planning process [consultation, informing, and placation, as categorised by Arstein (1969)]. At a minimum level it encourages the formation of partnerships between the municipality and the community in the delivery of new development and infrastructure, but it often goes beyond this by delegating power to citizens or enabling citizen control over LCI.

Of the cities studied only one, Freiburg, truly adopted the collaborative approach, although most encouraged some degree of public participation in the planning process. Freiburg adopted a collaborative approach towards planning new development and regeneration projects in the city which has led to the deployment of LCI. This is exemplified in the low-carbon district of Vauban.

To a large extent the residents and businesses in Vauban were in control of the development there. The planning authority established some environmental parameters for the development: the height of the buildings, a low-energy standard, a traffic concept, regulations concerning rainwater infiltration, and the greening of facades and roofs (Sperling, 2002). Within these parameters the community stakeholders made decisions about the development. They went beyond the municipality's parameters and expanded the environmental vision for the site. They (particularly the cobuilding groups) drove the inclusion of LCI [biomass district heating system, solar photovoltaic (PV) systems, and combined heat and power systems] in the neighbourhood. Some opted to operate and manage their own energy systems. Others adopted low-carbon living practices (for example, through the creation of car-share schemes and adoption of less parking provision). The community was engaged in visioning and design consultation exercises for the new urban district (Daseking, interview; (3) Sperling, 2002). This went far beyond the existing legal requirements in Germany for public participation in the planning process (Sperling, 2002). Three key groups were involved in planning Vauban: Project Group Vauban, (4) City Council Vauban Committee, (5) and Forum Vauban (6) (Brohmann et al, 2002) (see figure 1).

Workshops and exhibitions organised by Forum Vauban engaged all key stakeholders in thinking about and discussing the development of Vauban. It brought together citizens, architects, engineers, financial experts, and experienced managers of cobuilding projects (Daseking, interview;Sperling, 2002), and acted as the hub for the exchange of ideas, concerns, and expertise amongst these groups. It also helped to forge partnerships amongst stakeholders for delivering LCI and associated services.

In Vauban collaborative planning provided support for niche innovation originating within the community (Ries, personal communication; (7) Sperling, 2002). Community stakeholders determined the infrastructure integrated into the neighbourhood, which reflected their environmental goals, the local context, and their willingness to be involved in coprovision. This enabled the coevolution of low-carbon sociotechnical systems within Vauban.

The success of the collaborative approach in delivering LCI is highly dependent on context. Public subsidies, (18) supportive legislation, the local cultural context, ongoing political support for the collaborative approach and low-carbon systems, new institutional structures, and municipal ownership of resources all give greater leverage to collaborative planning and the adoption of LCI by the community, as demonstrated in Freiburg.

Various incentives can encourage the community stakeholders to adopt low-carbon systems. In Freiburg, federal and local subsidy schemes have been instrumental in encouraging the development of bottom-up, low-carbon initiatives. The energetic refurbishment of existing stock and construction of low-energy housing is subsidised by low-cost capital loans. (19) Capital grants are available for the installation of renewable energy technologies. (20) A feed-in tariff is available to the local community to invest in renewable energy projects. The regional energy company (Badenova) offers additional funds for local low-carbonenergy projects. Legislation has encouraged community stakeholders to adopt LCI. The feed-in law guaranteed grid access and remuneration for all renewable energy producers. Thus, renewable technologies became a secure and commercially viable investment. This law (18) Public subsidies, both capital and operational, for new low-energy buildings, energetic refurbishment, and the generation of low-carbon energy. (19) For example, KfW (Reconstructional Credit Institute) loans. (20) For example, the 100 000 solar roofs programme which subsidised the PV arrays on the football stadium. and the subsequent legislation have been central to community decisions to adopt LCI in Freiburg.

The cultural context influences community stakeholders' willingness to be involved in the collaborative planning process and how they prioritise environmental objectives. In the case of Freiburg, citizens and businesses are very keen to tackle their energy consumption and carbon footprint (Daseking, interview; Donn, presentation). Since the 1973 antinuclear protests the community has been supportive of low-carbon-energy alternatives. New environmental industries have sprung-up (21) (Lutzky, 2004), some focused on the production or operation of LCI.

Educational programmes have helped to develop local technical expertise for lowcarbon systems. Freiburg is now home to a large number of specialist energy institutions and enterprises. (22) It has become an innovation hub for renewable energy systems. Thus, community support for low-carbon industry, infrastructure, and services is significant. Hence LCI is integrated into the urban environment.

The collaborative planning approach is also in keeping with local cultural norms. The degree of community engagement in local decision-making processes and the provision of services in Freiburg as a whole are unusually high. The proliferation of community-led, smallscale renewable energy generation projects (23) and cobuilding groups in the city demonstrate this. Community stakeholders are keen to be involved in the codesign and provision of these new infrastructural systems.

This produces new institutional structures (eg, community energy cooperatives, community forums, and cobuilding groups) which can engage with public authorities in the processes of codesign and coprovision as well as provide the institutional frameworks through which community action becomes more effective. It has happened in Freiburg, but this degree of community engagement is unusual.

Long-term political support for collaborative planning is required for success. It takes time for the community stakeholders to develop the interest, expertise, and confidence required to engage fully in the process. Political support for LCI is also essential and has existed in Freiburg at least since the introduction of the solar plan in 1986. Municipal ownership of land has also been critical to the success of collaborative planning. It has enabled the city to designate spaces in which to test collaborative planning. It has also allowed community stakeholders to experiment with LCI and provided the opportunity for new institutions for coprovision to form. Overall it is the strength of the municipal leadership in Freiburg which has bought these elements together to deliver LCI.

It seems unlikely that this approach could bring about the same results in a different context. Nevertheless, the Freiburg case study does show that where stakeholders have the potential to be engaged in the planning and delivery of LCI, adopting this more selfdeterministic, bottom-up approach towards planning, in conjunction with supportive public funding and legislation, could help to facilitate a long-term transition towards low-carbon urbanism.

(21) In fact approximately 9400 people work in the environmental sector in the Freiburg Region, which constitutes 3% of all employees, worth €500 million per year (Lutzky, 2004). (22) Kiepenheuerinstitut für Solarenergie, Fraunhoferinstitut für Solare Energiesysteme, International Solar Energie Society, Öko-Institut and Solarfabrik. (23) Including local hydroelectric plants financed by housing associations; the solar collectors on Freiburg football stadium partially financed by the fans; and the small wind farm constructed on the edge of Freiburg, wholly financed by a community group.

In Freiburg collaborative planning encouraged community support for more innovative development models and LCI (Ries, personal communication; Sperling, 2002). It helped build the networks needed to support the long-term operation of new infrastructural systems. It built social capital amongst community stakeholders, enhancing the feasibility of collective approaches to the delivery of some services and utilities. Ultimately it resulted in a strengthened commitment to the ecological objectives of the plan amongst all key stakeholders (Ries, personal communication; Sperling, 2002). The range of community stakeholders involved in construction also created a diverse response to the low-carbon agenda. Enabling different groups within the community to choose their own focus and solutions ensured diversity and long-term support for new infrastructure (Daseking, interview; Ries, personal communication), most clearly demonstrated in Vauban. However, it also produced a more ad hoc approach to development.

Freiburg has adopted the collaborative approach across the city. So far this has demonstrated significant benefits in terms of building community support for both social and environmental agendas (Donn, presentation; Ries, personal communication). However, the collaborative planning approach appears to be extremely resource intensive, particularly for the municipality (Daseking, interview; Donn, presentation).

Involving a large number of stakeholders in consultation processes has inevitably lengthened development time lines and increased the cost of development (Donn, presentation). Thus, there are significant resource implications for planning authorities that are considering adopting the approach, and consequences for the speed of future development (Daseking, interview; Donn, presentation).

Finally, this approach creates 'protected spaces' in which communities can innovate and introduce LCI without having to compete directly with high-carbon infrastructure. It relies on communities to develop these niches of innovation, which can be a slower process than with systemic and market-shaping approaches where reliance is placed on industrial stakeholders to deliver.

Thus the proliferation of LCI can be slow if the collaborative approach is adopted. However, in the long term this approach may be more transformative. Certainly LCI is becoming more widespread in Freiburg, and the degree of public acceptance is growing but the process is slow.

The systemic approach

The systemic planning approach views the city as a complex ecosystem which metabolises resources through a variety of interdependent processes. The ultimate aim of this approach is to reduce the waste stream produced by the urban system by increasing the efficiency of urban processes, substituting renewable resources for finite resources, and promoting reuse and recycling.

The systemic planning approach reduces the environmental impact of cities by moving from predominantly 'linear' metabolisms to 'circular' metabolisms (Girardet, 2008). Its main advantages are that it can enhance the efficient use of resources (reduce CO₂ emissions), encourage greater energy self-sufficiency, and increase the resilience of cities.

The systemic approach tackles the creation of a low-carbon urban system in several ways. Carbon-reduction targets are integral to the formulation of any development strategy. They inform all decisions to introduce new infrastructure, expand cities, alter the land-use mix, increase densities, and alter locations in which activities happen. The planning system will support patterns of development (through the spatial plan) which lead to a reduction in emissions brought about by the introduction of efficiency measures and low-carbon energy sources. Urban planning authorities can also play a catalytic role in encouraging key stakeholders to deliver these integrated, circular systems. This is a top-down process. Planners create a vision to guide the production of the built environment in line with a city's environmental objectives. Of course this vision (in the form of a master plan) is subject to a public consultation process. The key stakeholders involved in the process of planning LCI systems for a city are the producers (developers, planners, architects, engineers, utilities, waste service companies, transport providers, and industry), whilst citizens play a passive role, acting largely as consumers. The approach is probably best exemplified by the cities of Stockholm and Freiburg. In some cities the spatial plan develops alongside energy and carbon reduction plans (eg, Freiburg and Stockholm). This ensures that planning authorities consider the energy and emissions impacts of the development decisions they make (Donn,presentation;(8) Stoll, interview (9) ).

In Freiburg development planning is one area targeted for the reduction of CO₂ emission (figure 2). The Freiburg Climate Action Plan informs the strategic plan and development decisions in the city. The planners analyse the impact of strategic land-use policies on CO₂ emissions as part of the process for creating the plan, using forecasting techniques (Donn, presentation). They also monitor the impact of the plan on CO₂ emissions after it has been applied, to determine which policies have been successful or unsuccessful (Donn, presentation).

A very similar process of target setting, forecasting, and monitoring has been adopted in Stockholm to inform spatial planning decisions (Stoll, interview). The Stockholm Action Plan for Climate and Energy sets the carbon reduction targets for the city. It identifies planned low-carbon projects and infrastructural systems, forecasting the relative impact of both on emissions.

The Stockholm Action Plan for Climate and Energy also informs the strategic planning process. Its objectives are a key consideration in the creation of the spatial plan and in strategic development decisions. The impact of these plans and policies on the reduction of CO₂ emissions is monitored by the municipality. The introduction of targets, forecasting, and monitoring processes has helped to deliver a step-change in how land-use planning tackles CO₂ emissions in Stockholm and Freiburg and led to the wider deployment of LCI. Systemic planning ensures that patterns of new development support LCI systems. This is well illustrated by the Swedish cities, Stockholm and Malmö, where high development densities and a mixture of uses have been maintained through the spatial plan to support district heating and public transport systems (Rosén interview; (10) Stoll, interview). Spatial factors will also influence the viability of more complex, circular, infrastructural systems: for example, the integrated closed-loop system in Hammarby (figure 3). However, a great deal more research is needed in this area to determine the optimal urban forms supporting a range of integrated, circular, infrastructural systems.

Urban planning authorities can also bring together the key stakeholders involved in delivering integrated LCI systems. The planning process provides a forum for a range of actors to discuss potential low-carbon systems and their integration. It also presents an opportunity for developers, utilities, and industry to form potential partnerships, which are shown to be essential for models of good practice in industrial ecology. Planners can set the parameters for development and encourage key stakeholders to work alongside each other throughout the development process to produce LCI systems. Unlike the collaborative approach, this is very much a top-down process.

In Malmö and Stockholm the vision to construct low-carbon neighbourhoods (Bo01 and Hammarby Sjöstad) with bespoke LCI systems was developed jointly by key stakeholders (11) through the planning process. The planning authorities established the baseline environmental goals for the neighbourhoods. This drove the inclusion of LCI in new developments. Key stakeholders were able to adopt their own technological approaches to achieving these targets, coordinated through the planning process.

The involvement of LCI providers in the creation of a spatial plan further supports the systemic planning approach. For example, in Freiburg, Badenova (the local energy provider) provides advice to the planning authority on low-carbon solutions for new development (figure 4), how new development can support existing LCI, and the potential for developing renewable energy generating capacity locally (Daseking, interview). This aids the deployment of LCI in the city.

The appropriateness of adopting a systemic planning approach towards the inclusion of LCI in cities, will also depend on the context. The political and institutional context, in particular, will influence its success. Municipal leadership has been critical for the successful delivery of LCI systems in Stockholm, Malmö, and Freiburg (Donn, presentation; Graham, interview; Stoll, interview). The municipality's role in each case has been to act as a catalyst, coordinator, and enabler of projects. Leadership from the municipality is essential for the success of systemic planning.

Changes in political control over time and clashes in political priorities at different scales can particularly affect the success of the approach (Donn, presentation; Stoll, interview; Vestbro, 2002). Political support for low-carbon development across regions and political leadership in cities are important for successful delivery. Political stability is also important, since political priorities are rapidly reflected in the planning system.

In Stockholm the changing emphasis placed on the environmental programme by successive municipal governments has greatly influenced development outcomes. When the principal decision to develop Hammarby was taken in 1995, Stockholm was governed by a red-green coalition, who supported an ambitious environmental programme (Vestbro, 2002). When the blue coalition took over in 1998 the environmental programme moved from being binding to a status of recommendation. This did not influence the integrated infrastructural system but it did affect the energy efficiency of stock connected to the system.

The blue coalition decided to sell municipally owned land to the private sector. The city authorities sold land on the Hammarby site to private housing and construction companies, and argued that land-lease contracts could include clauses about environmental standards, but in reality selling the land made implementation of an environmental programme much more difficult (Stoll, interview;Vestbro, 2002).

A clash in the political contexts at a city-level and regional scale can also influence the potency of a systemic planning approach. The political clash between the city of Freiburg (historically supportive of environmental priorities and specifically renewable energy) and the region of Baden-Württemberg (historically conservative and supportive of nuclear energy) illustrates this. A difference in political priorities led the regional and local planning bodies to work in opposition in terms of delivering renewable energy (Donn, presentation).

The regional planning body prevented the development of wind farms in the region surrounding Freiburg, whilst the municipal planning authority positively promoted all forms of renewable energy within the city boundaries. This limited the options for the development of renewable energy capacity to within the city limits of Freiburg (only about 30 megawatts of installed energy) and resulted in the focus being on biofuel and solar systems. It also limited the potential for LCI.

The institutional context is also crucial. The power of the municipality in particular can greatly influence the potency of a systemic planning approach. Municipal control over resources-land, funds, utilities, and housing companies-strengthens the approach. For example, the municipal ownership of sites in Freiburg (eg, Vauban and Reiselfield), Malmö (eg, Bo01), and Stockholm (eg, Hammarby and Royal Stockholm Seaport) meant that the planning authorities had greater control and could impose environmental programmes on new development. In fact, they required the introduction of LCI in return for the release of land. Further leverage was given for the integration of LCI in Swedish cities through significant public investment in site improvements, the development process, and infrastructure (Graham, interview; Stoll, interview).

Municipal ownership of housing companies, utilities, and waste and transport services can also ensure the deployment of LCI in cities. In Stockholm, it was the close collaboration between the municipally owned utilities (energy, water, and waste) and the urban planning department which resulted in the development of the prototype integrated closed-loop system. The municipality acted as the multi-interest institution coordinating development and LCI in this instance. Since privatisation of the energy system (after Stockholm Energi was taken over by Fortum Energy in 2002), coordination of infrastructural systems and development has become more difficult (Stoll, interview).

A systemic approach really requires that multi-interest institutions implement and operate integrated LCI. Municipalities work well in this role. Private utilities tend to be singleinterest institutions, unconcerned with promoting integrated infrastructural prototypes. The involvement of various institutions with different cultures, structures, and organisational processes in the creation of integrated low-carbon systems can make working together difficult (Jonsson, 2000;Pandis and Brandt, 2011).

Freiburg provides an alternative model to Stockholm. In Freiburg single-interest stakeholders (from the public and private sector) work together as a municipal steering committee for climate protection. They are actively involved in deciding on how to deliver carbon-reduction targets and LCI in the city. This coordinates the actions of stakeholders and ensures they have some vested interest in delivering infrastructure compatible with these goals. In addition the municipality owns 33% of the shares of Badenova, which enables it to have significant input into the company's decision-making processes. This also helps to reinforce the working relationship between the city planning department and the energy company in the production of spatial, energy, and climate plans.

Finally, the focus of systemic planning is on transforming industrial regimes (in this instance the construction, energy, waste, water, and transport industries in particular), their products and systems. It requires that industry is prepared to innovate and not be locked-in to the existing culture, structure, and practices. Without industries that are willing to innovate difficulties in implementation may arise. Thus, industrial innovators are critical to the success of the systemic planning approach. Of course, the emergence of innovators can be encouraged by the context, for example, through regulation and fiscal incentives.

The systemic approach creates a protected space in which industry can test innovative LCI (eg, Hammarby, Bo01, and Vauban). Industrial stakeholders have greater capacity (supply chains, skills, and financial and human capital) to innovate and produce low-carbon systems at scale. They benefit from their involvement in this process, as developing new infrastructural systems provides them with a market advantage. For example, SWECO (the team who developed the closed-loop system in Hammarby) have gone on to plan similar systems for other cities worldwide.

This approach provides clear guidelines (and support) for those involved in the delivery of LCI, and adequate flexibility for innovation. It engages industrial stakeholders in the process of establishing the guidelines, which provides certainty and reduces risk for those investing in new systems (eg, Bo01 and Hammarby). The transition is systemic (rather than ad hoc) and can occur rapidly if given adequate political, financial, and regulatory support. It enables the creation of integrated systems rather than a series of separate systems. This creates potential for greater synergies between systems and resource efficiencies.

Integrated infrastructure systems can offer unaltered service dimensions of volume, content, and quality but lessen environmental impact by reducing the overall service volume (Jonsson, 2010). They can significantly reduce infrastructure and CO₂ emissions by reductions in material and energy use, especially when integrated with renewable-energy sources (World Bank, 2010).

An evaluation of Hammarby compared with a baseline scenario showed a 29%-37% reduction in CO₂ emissions (Brick, 2008). Integrated infrastructure systems produced by systemic planning also tend to be more resilient, as their localised, circular, and renewableenergy-supported metabolisms enable them to be largely self-sufficient, reducing vulnerability to external disturbances (Hodson and Marvin, 2011).

However, the approach does not necessarily alter the behaviour of communities. In both Malmö and Stockholm community engagement has played a very marginal role in the adoption of LCI. Thus, the local communities' energy awareness and behaviour has not altered. The social and technical systems have not coevolved (in contrast to the situation in Freiburg). The communities have not been empowered to become coproviders or encouraged to adopt more energy-efficient behaviours.

Nevertheless, the systems adopted in Bo01 and Hammarby have produced energy and carbon savings without the engagement of citizens. Thus, passive systems created through a systemic approach can deliver CO₂ savings even without community engagement. However, the savings made tend to be less than predicted (Brick, 2008;Nilsson and Elmroth, 2005). This is for three key reasons: poor assembly of systems, inaccurate predictive models, and human behaviour (Stoll,interview;Svenberg,interview (24) ). The systemic approach does not help to build capacity or willingness within communities to tackle carbon reduction.

The implications of systems integration are unknown, largely due to a lack of data (Pandis and Brandt, 2011). However, there are some concerns that the integration of systems might create barriers for more fundamental infrastructural changes in the future (Pandis and Brandt, 2011). For example, integration could create lock-in for specific technical systems that may become redundant due to changes in climate or scarcity of resources. Environmental, cultural, institutional, economic, and political changes may also result in the redundancy of systems.

Planning the integration of these complex systems is difficult. It requires planners to act as intermediaries between key stakeholders, mediating between their different priorities in production and realisation phases (Guy et al, 2011;Hodson and Marvin, 2011). This will lengthen the development process. The systemic approach will also require improved monitoring processes to determine the effectiveness of policies supporting infrastructural provision and the effectiveness of the systems themselves (Donn, presentation). This increases the role of planners significantly and in all probability will lengthen and increase the cost of the planning process.

The market-shaping approach

The market-shaping approach relies on developers, utility companies, energy service companies, and multiutility service companies to deliver low-carbon urban systems. Unlike the systems approach, the market-shaping approach lacks a municipally led master plan for tackling carbon reduction. Instead markets for LCI are shaped by planning-based incentives.

The market-shaping approach uses the regulatory function of planning (length of planning process, design codes, community infrastructure levies, and access ordinances) to encourage the private sector to fund, build, and operate LCI. The approach is used to reduce development costs, investment risk, and/or provide a source of funding to assist in the development of LCI. The presumption is that once LCI is integrated into the built environment demand will grow for the resources and services they offer, thus shaping the market.

The various planning-based mechanisms used to shape the markets for LCI cannot be exemplified by a single case-study city. San Diego, Boulder, London, and Freiburg demonstrate the breadth of mechanisms being adopted, whilst Malmö provides an illustration of how relaxation of planning controls can lead to the perpetuation of a high-carbon infrastructure in cities.

The length of the planning process increases the net build cost of low-carbon development. The longer the planning process, the more costly it is for the developer. Thus, reducing the length of the planning process can provide a significant financial incentive to a developer to build a low-carbon development (Navigant Consulting, 2008).

In the city of San Diego planning permits are fast-tracked for developments with a substantial PV system. (12) Permitting periods have reduced from 24 months to 6-9 months (Farhar et al, 2004). This has saved developers the finance charges on loans and covered the costs of the PV systems installed (Farhar,interview;(13) Farhar et al, 2004). The policy has also led to an increase in applications for solar houses in San Diego County and is being piloted elsewhere in California as a method to encourage deployment of PV technology (Navigant Consulting, 2008).

Planners can use standards and codes to encourage the deployment of LCI. Both provide developers with clear objectives for construction, thus reducing the risk of investment in expensive LCI. Introducing standards at a local level enables their practical application to be tested. It also provides industrial stakeholders with a protected space in which to develop compliant low-carbon systems (supply chains, expertise, and systems of production), which could not otherwise compete with high-carbon alternatives.

Site-specific energy-efficiency standards have been used successfully by planning authorities in various European cities (eg, Stockholm, Freiburg, and Malmö) to promote the deployment of low-carbon systems (Donn,presentation;Graham,interview;(14) Stoll, interview). Subsequently, these standards have been adopted city-wide (eg, Freiburg, Hannover) and in Germany they have also informed the federal energy-efficiency standard, which has started to generate demand for energy-efficient buildings in Germany.

In the London Borough of Merton, a standard for the introduction of on-site renewable energy for all new developments over a minimum threshold size was introduced in 2003. This standard was subsequently adopted by many local planning authorities in London (as well as the Greater London Authority), the Thames Gateway, and across the UK (although modified in some instances). The Merton rule (as it became known) demonstrated some 'patchy' success in terms of delivering LCI for new developments (Williams, 2010). In the areas where it operated, the Merton rule provided developers with greater certainty, encouraged an increase in supply, and began to generate public interest in LCI in urban areas.

A more strategic approach to the delivery of LCI is now being adopted in London. The community infrastructure levy (CIL)-a development tax-will be applied to all new development in London by 2014. It will be used to provide subsidy for much needed infrastructure, (15) including LCI, and to fund the operation of existing infrastructural systems. The percentage of the CIL spent on LCI will vary depending on the local authorities' priorities. The charges levied will also vary geographically and with land use.

Investment in LCI in London at a strategic and local level will be guided by a decentralised energy master plan created by the London Development Agency. This in turn will inform the local development frameworks in each borough, which will guide spending on LCI and (12) The PV system must provide 50% of the estimated electricity needs of the building or development. the levy charged. Thus, the availability of funding (provided by CIL) and greater certainty for industry (provided by the energy master plan) should result in the wider deployment of LCI in London, although this will depend on local priorities and leadership. Improved public access should also help to raise awareness and potentially create future demand for LCI (ICARO Consulting, 2009).

Planning can also ensure the long-term integrity of LCI alongside new development. The threat of new development impoverishing access to infrastructure greatly reduces potential market demand for new systems and producers' willingness to provide them. The city of Boulder has implemented a solar access ordinance which guarantees access to sunlight for homeowners and renters in the city. The ordinance sets limits on the amount of shading permitted by new construction. The city is zoned into three different areas, each with different solar access requirements dependent on existing urban form. The ordinance provides those investing in new solar systems with assurance of their continued ability to operate them effectively. This is fundamentally important to the supply and shaping markets for solar technologies in Boulder (Merrigan, interview; (16) Reed, interview (17) ). Access design codes could also be developed for other LCI systems to optimise their operation and ensure long-term design integrity, thus protecting markets.

The approach works well in contexts where private investment is the main source of funding for the built environment, and government intervention is limited. The codes, standards, strategic energy plans, and access ordinances provide guidelines for those developing urban areas. They provide certainty for those building new low-carbon districts and installing infrastructure. They also offer protection to those investing in existing low-carbon systems. This reduces risk for investors. Political stability and a long-term policy for shaping markets are required in order to provide further certainty for investors. In the UK constantly changing renewable energy policy has led to slow deployment of LCI (Lipp, 2007;Painuly, 2001;Ragwitz and Held, 2006;Reiche, 2002).

The market-shaping approach is suited to an institutional context in which the planning system is less interventionist (as in the UK and USA). In this context, the role of the planning system is to support certain forms of development (shape markets), as well as to encourage industrial stakeholders to innovate. The planning system creates 'protected spaces' in which industrial stakeholders can innovate. It does this by reducing the short-term and long-term costs to industrial stakeholders of delivering new technical systems.

Energy service companies and edge-of-service providers can be critical to the operation and market acceptance of new infrastructural systems (Reed, interview). They ease the transition by reducing consumer involvement in the installation, management, maintenance, and operation of new systems. This in turn increases market demand. New institutions for supporting and servicing LCI may emerge, as demonstrated in San Diego (with the introduction of renewable technologies). However, where they are slow to emerge deployment will be limited.

Financial incentives for industry also support the market-shaping planning approach. For example, capital or operational subsidies (eg, feed-in tariffs and capital grants for PV cells in Germany and California), or access to low-cost municipally owned land (eg, Vauban, Hammarby, and Bo01) can provide greater encouragement for industry to innovate. Market transformation programmes have also been instrumental in market shaping as demonstrated by the programme adopted in San Diego (Navigant Consulting, 2008). However, the longterm success of this approach relies on a market for LCI eventually developing. Thus, the cultural context is also very important.

Planning controls can encourage the deployment of LCI before it becomes commercially viable. However, with a market approach these controls are eventually removed (alongside other market incentives) as demand grows and LCI is integrated into mainstream development models. In the absence of public demand, longer term, interventionist approaches are required to ensure industrial stakeholders continue to integrate LCI into new development.

Long-term support for the market-shaping approach is also needed. Even with the marketshaping approach some degree of municipal leadership is required to ensure that standards, codes, ordinances, and development taxes are applied systematically, and that revenue raised is invested in LCI. Removing support for the approach is likely to result in failure. For example, in Malmö planners did not impose the same strict environmental guidelines on the second phase of the Bo01 development as was used for phase one. Thus, phase 2 achieved lower building-energy-efficiency standards (Graham, interview). In addition the low-carbonenergy system constructed in phase one was not linked to phase two of the development (Rosén, interview). According to those involved in the project, there was no demand for LCI and thus it was too early to remove the mandatory environmental standards set for phase one.

The market-shaping approach also encourages innovation. It creates a protected space in which new infrastructural systems are tested before markets develop (eg, solar homes in San Diego and Boulder). It can help innovative industries achieve a market advantage, particularly where landscape changes to legislation (25) or economic trends (26) support it. This may lead eventually to the transformation of industrial regimes, although there is no evidence of this happening amongst the case studies.

Market-shaping relies less heavily on public subsidy for new infrastructural systems, as the majority of funding comes from the private sector. The willingness of the private sector to invest in infrastructure cannot be guaranteed or enforced through the planning system. Inertia to change within industries created by culture, structure, and practices, creates barriers to the delivery of LCI (Williams, 2011). Even if planning regulations reduce building costs and investment risk, industrial lock-in creates huge inertia to change. It is unlikely that planning controls alone can overcome this industrial inertia (Williams, 2008;.

In instances where the market-shaping approach has been effective additional subsidies have been used to encourage growth in demand, or mandatory building-energy standards have been used to encourage growth in supply. For example, in San Diego operational subsidies (a feed-in tariff) generated demand for buildings with PV, whilst the fast-track planning system in combination with capital subsidies (federal tax credits) for renewable energy encouraged the construction industry to increase supply. It was the combination of these incentives which encouraged market growth in California (Navigant Consulting, 2009).

An advantage of this approach is that it offers flexibility. Low-carbon standards and solutions can be tailored by planners to the local context and current technologies. However, this degree of flexibility can also reduce certainty for industry, which limits willingness to invest in new systems. It results in the ad hoc deployment of infrastructure. For example, the variation in the application of the Merton rule amongst planning authorities in Greater London and the Thames Gateway limited the construction industry's willingness to invest in decentralised renewable energy systems in the region (Day et al, 2009;Williams, 2010).

The Merton rule example demonstrated that there tends to be great variation in interpretation and implementation of planning objectives between localities, which slows the deployment of LCI (Williams, 2010). Different municipalities have different priorities. Carbon reduction is one of many competing local targets which are weighed against each other in development decisions.

Local variation in technical expertise amongst planners also influences interpretation of guidance. Lack of guidance, unclear guidance, or complicated codes may also affect interpretation and implementation of planning controls (Williams, 2008). Planning can only be a useful tool for creating certainty for industry if it is applied consistently. Otherwise, it is more likely to create greater confusion within industry. Other regulatory tools are likely to produce more-consistent results (Williams, 2008;.

The planning process itself can be shortened; equally it can take longer to appraise a development which integrates new infrastructural systems. As planners' expertise in assessing planning applications increases, there is likely to be a reduction in the length of the process. For example, as expertise grew the application process in London for developments containing decentralised renewable energy systems reduced from 700 days in 2004 to 100 days by the end of 2005 (Day et al, 2009). Similarly the resources needed to assess more-complex applications will reduce over time, as LCI becomes part of the mainstream model for development.

This approach is unlikely to encourage the coevolution of low-carbon sociotechnical systems or coprovision. It is an approach focused on easing the adoption of technologies by industrial stakeholders rather than by the wider community. However, the integration of LCI into new development enables the community to test new infrastructure. Over the long term this will also help to develop markets, as demonstrated by solar homes in San Diego.

An expended typology

An analysis of the case studies identified three planning approaches to the delivery of LCI in cities. A summary of the new approaches is provided in table 2. The goal of the approach, the role of planning in delivery, the specific planning mechanisms utilised, the nature of the planning process (top-down and bottom-up), and the key stakeholders involved in implementation are the dimensions used to characterise each approach. These approaches are subtly different from those presented in section 2. The collaborative approach goes beyond involving stakeholders (community, industry, and municipality) in the visioning, designing, and planning of LCI. It encourages stakeholders to own, operate, and manage LCI. This greater level of investment in LCI increases long-term community support for projects.

The systemic planning approach goes beyond the strategic coordination of land use, urban form, and LCI, in advocating a systems approach towards the low-carbon planning of cities. It recognises the interdependencies and intradependencies between processes, infrastructural systems, and stakeholders, and it takes a more holistic approach towards coordinating infrastructural systems, plans, policies, and stakeholder input. It also establishes a method for determining the impact on carbon emissions of changes to the system as a whole and in parts.

The market-shaping approach goes beyond merely regulating the supply of LCI. It also encourages growth in supply and the formation of a market for LCI, achieving this by using the planning system to increase the cost-effectiveness, reduce investment risk, and increase the efficiency and longevity of LCI.

The context

The context (economic, cultural, regulatory, institutional, political, technological, and environmental) in which a planning system operates will influence the viability and effectiveness of the three approaches described.

The influence of context

The analysis of the case studies identifies a number of contextual factors which provide greater leverage for the successful implementation of the three planning approaches. These are summarised in table 3.

For all three approaches political stability and long-term policy goals, supportive institutional structures, municipal ownership of resources, municipal leadership, public subsidies, and supportive regulation are important for successful application of LCI. For the collaborative approach to be successful the presence of community innovators in the local context is also crucial. The systemic and market approaches rely more on industrial innovators. The systemic approach requires strong municipal leadership and support for the introduction of LCI and effective coordination of key stakeholders in the delivery of integrated low-carbon systems. For the market-shaping approach to be successful long term demand for LCI needs to be established. This will be influenced by the economic, cultural, and institutional context in which systems are embedded.

Advantages and disadvantages

Each planning approach has its advantages and disadvantages. A brief analysis is presented here.

Summarising the advantages and disadvantages of the three approaches

All three approaches can potentially enable innovation by providing protected spaces in which LCI does not have to compete with high-carbon infrastructure (table 4). If applied systematically all three approaches could increase certainty for investors. The collaborative approach has the added benefit of encouraging social and technical systems to evolve together. This will produce long-term community support for LCI, as well as increasing community energy awareness and promoting behavioural change essential for the long-term success of LCI. The market-shaping approach's key benefit is that it should help to create markets for LCI, thus limiting the need for public subsidy; whilst the systemic approach should result in wider coverage of LCI in cities.

Initially all three approaches are likely to be resource intensive and increase the length of the planning process. This is probably the greatest weakness of the collaborative approach. In the case of the market-shaping approach the length of the process will reduce once capacity is built in planning departments to assess the potential for LCI. Generally, the market-shaping approach should increase certainty amongst investors; however variation in application can undermine this. A major problem with the systemic approach is the passive role taken by the community, which limits community support for LCI and the potential for real behavioural change. The key shortcoming of the market-shaping approach is its reliance on private investment for the introduction of LCI (table 5).

Increases risk and uncertainty for investors

✓ b Relies on creation of markets ✓ a The market-shaping approach can lengthen and shorten the planning process. b Variation in application can increase uncertainty for investors.

Conclusions

This analysis suggests that a variety of planning approaches can be used to encourage the introduction of LCI into cities (table 3). The typology developed here is not exhaustive, it merely demonstrates the planning approaches adopted in cities where LCI has been integrated successfully. Unlike previous research it identifies a typology which is specific to the introduction of LCI, rather than the wider objective of encouraging CO 2 reduction in cities. It also illustrates, using a range of international case studies, the diversity of approaches that could be adopted. However, it should be noted that the political economy and regulatory and institutional framework in which each planning system is embedded varies between countries. This may influence the extent to which the planning approaches outlined in this paper are transferable without wider changes within the system in which they are embedded. The success of each planning approach will be context dependent (table 4). For example, the systemic approach works well in cities where the municipality has substantial powers, as demonstrated by Freiburg and Stockholm. The market-shaping approach is more suited to cities where municipal controls and resources are limited but industries may be more willing to innovate, as demonstrated in San Diego and Boulder. The collaborative approach will only be successful in delivering LCI in locations where citizens are proactive and proenvironmental.

It is difficult to dissociate the impact of a particular planning approach on the delivery of LCI from other regulatory, educational, and fiscal instruments. For example, in Freiburg local and national subsidies for low-carbon technologies in combination with the ecological tax and renewable energy regulation have played an instrumental role in encouraging citizens to adopt LCI. Similarly, in San Diego subsidies and market transformation programmes have incentivised the development of solar homes and stimulated demand.

In Stockholm, the introduction of LCI has been heavily subsidised by the state. Alongside this the introduction of a carbon tax and subsidies for biofuel has encouraged systemic shifts in attitudes within the city. In all cases the planning system alone is not responsible for the adoption of LCI. Combinations of fiscal, regulatory, and educational tools are needed to provide leverage.

The analysis has also demonstrated that approaches are not mutually exclusive. For example, in Freiburg the collaborative, systemic, and market-shaping approaches have in fact operated together very effectively to deliver LCI. Here the local context supports all three planning approaches (which is unusual). The effectiveness of planning as a mechanism for delivering LCI appears to be greatly enhanced by this combination approach. The best explanation for this is that the combination approach encourages all key stakeholders (community, municipality, and industry) to support and deliver LCI in cities.

The evidence from the examples suggests that planning can be used to provide 'protected spaces' in which industry, citizens, and municipalities (in various combinations) can develop low-carbon sociotechnical systems. In these spaces low-carbon systems no longer have to compete with high-carbon alternatives, which are the default option due to industrial and cultural inertia. The whole city or even a city region may act as a protected space (eg, Freiburg and Stockholm) or alternatively areas within cities may offer protected spaces (eg, London).

This phenomenon can be seen in Germany, where federal regulatory and fiscal incentives have been introduced to encourage the adoption of LCI. However, only a few German cities have introduced LCI (eg, Freiburg, Hannover, Munich, and Hamburg). In all cases municipal leadership has been instrumental in the delivery and the planning system has played a key role in creating protected spaces in which LCI could be adopted.

Equally Greater London and the Thames Gateway region demonstrate the importance of the planning system in the delivery of LCI. The variation in application of the Merton rule across the region has greatly influenced the adoption of LCI, even though national fiscal and regulatory mechanisms to encourage its introduction remain constant. Again, a combination of municipal leadership and a proactive planning system have been instrumental in providing protected spaces in which LCI can flourish.

The question is: which approaches are likely to be most successful for delivering LCI in the future? The collaborative and market-shaping approaches perhaps fit better with the emerging entrepreneurial and collaborative forms of governance [as described by Healey (1997) and Harvey (1989)]. Yet, the systemic approach appears to have the greatest capacity for generating a more rapid transition to low-carbon systems that is needed if we are to combat the resource crisis and climate change. Thus, success will not only be a function of the context in which planning operates; it will also be a function of cities' overall objectives.