Emerging forms of hybrid energy systems in cities of the global South: Summary of the Delhi Roundtable

Summary of Roundtable on « Emerging forms of hybrid energy systems in cities of the global South »
Centre for Policy Research, New Delhi, 30th October 2018

Summary by Federico de Lorenzo and Bérénice Girard

Videos of the presentations are inserted below after the summary of each panel

Organizers and moderators: Marie-Hélène Zérah (IRD and Centre for Policy Research), Ankit Bhardwaj (Centre for Policy Research), Federico De Lorenzo (Sciences Po Paris), Augustin Delisle (Agence Française de Développement)
Speakers : Eric Verdeil (Centre for International Studies, Sciences Po Paris), Rémi de Bercegol (PRODIG, CNRS, Paris), Ankit Bhardwaj (Centre for Policy Research, New Delhi), Gaurang Sethi (Azure Power, India), Santosh Kumar Thakur (Energy Efficiency Services Limited, India)

Organized jointly by the Agence Française de Développement (AFD), the Centre for Policy Research (CPR) and the Centre de Recherches Internationales (CERI, Sciences Po) and hosted by the CPR, this roundtable brought together around 30 stakeholders from urban studies and the energy sector to discuss the coevolution of electricity supply systems and urban change in the Global South and understand how the imperative of energy transition translates in cities. The roundtable gathered researchers and academics from Indian and French institutions (such as TISS, CEER, CNRS), representatives from NGOs (Centre for Science and Environment, Shakti Sustainable Energy Foundation), private and public companies (such as Oracle Utilities, Maxop Engineering, India Smart Grid Forum), and international organizations (such as World Bank, International Solar Alliance, European Investment Bank and the German Developement Bank -KfW).

The roundtable started with a presentation of the ANR funded Hybridelec project by É. Verdeil. As cities of the Global South are experiencing fast economic and population growth, their electricity supply systems often lag behind (cuts, load-shedding, high prices, inadequate coverage…). In this context, public policy has mostly been focused on extending the grid and augmenting supply. However, this development is often not going fast enough and alternative responses emerge both in cities and in the intermediate spaces between them (generators, solar panels, batteries – different technologies bringing new opportunities and often transformed to suit the local needs). In this context, the Hybridelec project aims at thinking together local electric supply systems and urban change by focusing more specifically on the collective solutions, often socially more structuring and technically more complex, than conventional grid provision. How are these alternatives initiated, organized and regulated? How do they sustain and extend? What are the implications for the electrical system and the objectives of energy transition? Making the hypothesis of an increased heterogeneity of these systems rather than that of a convergence, the Hybridelec project wishes to go beyond the idea of delivery palliatives and uses the concept of hybridization to look at the long-lasting process of electrification and to understand the complex assemblages of actors, technical objects and institutions contributing to changing energy systems on the ground.
R. de Bercegol then presented his work on slum electrification in Nairobi, Kenya. Though, officially, the Kibera slum does not have electricity (due to unidentified land tenure), it is electrified, as cartels provide electricity by poaching it from the conventional network. In 2014, a new electrification project was launched in Kibera by the Kenya Power Corporation, justified by the publicly-declared goal of 100% electrification by 2030 and the technical, security and financial problems the illegal poaching leads to. Though there have been tensions in the past between engineers and slum dwellers, the situation seems to have improved in recent years as the power corporation resorted to different means of ensuring the sustainability of the grid connections: involvement of cartel members in the programme ; installation of prepaid and locked meters to avoid unpaid bills and tempering ; use of aluminium instead of copper for the wires to avoid theft, etc. R. de Bercegol then turned to the societal meaning of slum electrification, by showing how prepaid meters act as a depoliticization tool in a long-lasting power struggle, but also as an instrument to formalize poor consumers. Indeed, though they take part in the regularization of slum dwellers by acknowledging their right to live in the city, they also act as disciplining and punishing tools for the poorest consumers, who are not able to afford the conventional network’s electricity. The paradoxical consequences of the arrival of the grid were also underlined, such as changes in daily practices (households going back to charcoal to cook), rapid tempering of these same meters and continued predominance of the informal system of provision.

The two papers were followed by a rich discussion which mainly focused on how to reconcile two policy paradigms often seen as contradictory, namely centralized versus decentralized energy production. Several participants called for a change of perspective, as mini-grids are often seen by distribution companies as sources of potential loss of revenue and influence, and renewable energies are mostly developed in a centralized manner. In this context, how can we ensure a change of perspective from “centralized or decentralized” to “centralized and decentralized”? Taking the case of Beirut, É. Verdeil underlined how, as solutions first seen as temporary become more and more embedded in the conventional network, we observe a shift in representations amongst NGOs, international funding agencies and even ministries. Only the national or regional utilities in charge of production and distribution are still reluctant to promote these systems. However, one can hope that as more sustainable technologies develop on the ground these same utilities will get on board. Innovative solutions to provide electricity to Kibera were also discussed, such as user group connections, solar panels bought in bulk or batteries.
Based on a joint ongoing research with F. De Lorenzo and M.H. Zérah, A. Bhardwaj’s presentation focused on Indian Smart Cities. Even if energy is not generally framed as an urban issue, the Smart Cities Mission (SCM) seems to leave scope for cities to actively shape their energy systems. Findings from the analysis of the Smart Cities proposals offer some interesting insight in what energy management in the Indian Smart City would look like. Although in the proposals cities focused mostly on basic energy infrastructures (expansion of the power network, increasing energy production, etc.), there is considerable room for low-carbon technological improvement as well. In particular, the Smart Cities will direct their efforts towards four low-carbon technologies: solar (mostly rooftop), energy-efficient street lighting, E-mobility, and waste-to-energy. Smart meters receive surprisingly small investment. However, notwithstanding the significant budget planned for the four sectors overall, most of the cities do not seem to have envisioned a global low-carbon strategy, but rather each city seems to have selected some specific area(s) of action.
Head of Business and Project Development at Azure Power, G. Sethi presented the potential and challenges of the solar rooftop sector in India from a private company perspective. He first underlined the current favourable context for solar rooftop development in India: an ambitious programme launched by the Ministry for New and Renewable Energy for the installation of 40 GW of solar projects by 2022; an estimated 124GW solar rooftop potential in India; a supportive regulation and policy framework and a growing demand, mostly coming from the government sector but also from the industrial, commercial and, to a lesser extent, from the residential sectors. He underlined the important role the government sector plays as an initiator in the market, as well as the reduction in solar power tariffs over the last few years. He then turned to challenges the sector is currently facing: a focus which remains mostly on large solar utility; limitations regarding capacity installation, as discoms do not wish to see building owners become producers of energy; a lack of coordination between Central and State governments ; different requirements and regulations from State to State, but also the existence of unfavourable charges for supply through open access and the small number of tenders released by discoms to buy solar-produced energy.
S. K. Thakur offered an overview of EESL’s action as its General Manager of the street lighting division. EESL is a joint venture of four public companies which facilitates and provides support for the implementation of energy-efficient lighting projects in India. Availability of funding to start LED projects is the main challenge that urban local bodies are faced with. EESL can take on the initial investment for them. Moreover, it has been able to bring prices of LED down dramatically through demand aggregation and bulk procurement. The Street Lighting National Programme (SLNP) is one of the most successful initiatives undertook by EESL. Its target is to replace 13.4 million street lights with LED lighting in Indian towns and save 1,500 MW of energy. Drawing on its experience, EESL has adapted its business model to specific capacities of the urban local bodies and technical constraints. The agreement between EESL and local bodies includes support by the former for supply, installation, and maintenance (tasks delegated to different agencies) of the street lighting assets, other than financement. Poor infrastructures and repayment capacity of urban bodies are amongst the most common challenges the SLNP is faced with.

Throughout the second Q&A session, the debate focused on the role of urban authorities in shaping energy transitions in India, on the importance of access to accurate data, and on standards of quality of technologies. First, the view that the SCM represents an opportunity for local bodies to be leading actors of the urban energy transition, is challenged by a first-hand insight in the solar rooftop sector which shows that growth here has been primarily driven by tendering opportunities released by State and Central authorities. Second, data on the technical features of the places of transition are necessary both for authorities to set realistic targets and for implementing bodies. In particular, collection of reliable data on government buildings, along with the necessity to obtain several authorisations from the concerned bodies, represents a crucial question for Azure and the private sector in general. The complexity which Azure is faced with in implementation contrasts with EESL’s less-stringent requirement to have its agreement signed by the sole municipal commissioner. Last, while EESL’s primary concern is to expand the lighting network than meeting quality standards, in a PPA/RESCO business model (used by Azure), it is in the best interest of the developer to deliver high-quality technologies because they own the assets.

In conclusion, this roundtable showed the diversity and complexity of the possible pathways to energy transition in urban contexts. In the case of India, though use of REs is growing fast thanks to ambitious national schemes, decentralized energy production systems are facing many challenges such as a lack of capacity at the local and regional levels and paradoxical regulations and policies at the different stages of decision-making. The roundtable thus confirmed the necessity of in-depth fieldwork and continued discussions between policy-makers, stakeholders and academics to confront the mismatches between the declared ambitious objectives of energy transition and the local and technical constraints.

Observations on some hybrid energy systems in Delhi and its periphery

The Hybridelec research project started to explore the Indian reality, inparticular the city of Delhi and its southern suburb, during group visits which were conducted in the wake of the seminar co-organised by AFD, CPR and CERI on 30 October at CPR (see separate report). These field trips (Marie-Hélène Zérah,Bérénice Girard, Rémi de Bercegol, Gowda Shankare, Federico de Lorenzo, andEric Verdeil) have allowed us to identify five socio-technical setups and observe their typical ways of producing and accessing electricity.

Solar-roofed public buildings


Photovoltaic plant on the roof of NPC, Delhi (Eric Verdeil, CC)

The first setup concerns numerous public agencies and enterprises which are having solar equipment installed on their roofs. For example, in the government district of south Delhi, we visited the National Productivity Council (NPC), a public enterprise acting as a consultant which helps public agencies or enterprises to improve their productivity, including with regard to their environmental or energy policies. When we arrived, there was a power cut which, according to our hosts, was highly unusual. Nevertheless, a dedicated generator powered the lift and emergency lighting.


Other plants on the roofs of neighbouring buildings in the government district of Delhi (Eric Verdeil, CC)

Those in charge of NPC’s Environmental Management Department showed us a system of photovoltaic panels, with a total capacity of 30 kW, which had been installed on their roof in 2015. The system was connected to thehigh-voltage grid it supplied. Thus, the electricity produced was not directlyused in their building, however, NPC benefited from a reduction in itselectricity bill equivalent to the number of kW injected into the (high-voltage) grid. The electricity generated represented around 10% of thebuilding’s annual consumption (Rs. 30,000 per month) so this was a saving for the enterprise. Our interlocutors strongly emphasised the need for regular cleaning to remove the dust and maintain good production capacity. Several other buildings in the surrounding area were also equipped with solar panels of thesame type, albeit they were a lot less well maintained. Although this was not very clear, our interlocutors put forward several contributing factors behind this recent equipment:

  • a decision by the central government that tends to oblige public agencies to have this kind of systems installed;
  • the adoption of rules that favour net metering (which, however, did not apply in this case since the electricity produced was fed directly into the HV grid);
  • the project was completed with the help of a grant obtained from the Ministry of New and Renewable Energy as part of the National Solar Mission initiative;  
  • another incentive for this project was the obligation to have solar panels installed in order to be considered a “green building”;
  • the financial incentives were substantial overall: the project, which cost a total of Rs 2,157,000, received Rs 647,000 in subsidies;
  • finally, a swift drop in the cost of solar panels which has halved since 2015 was another essential factor in this investment.

The policy of Delhi Metro Rail Corporation (DMRC) for installing solar panels

The DMRC represents a second very impressive example of solar-panel installation. Wevisited a workshop depot east of the Yamuna River, in the Karampura district. Our interlocutor showed us a 100 kW plant installed on the roof but, in reality, all the roofs of the complex, including round-shaped ones which arenot very suitable, as well as, more widely, metro stations and various empty fields were covered in panels. This was the result of a series of contracts since 2015, with an installed capacity reaching 24 MW in 2018, and a new installation phase was currently under way. DMRC’s objective is to produce a part of the electricity it needs for traction and lighting. We learnt that solar production represented 30% of the daytime consumption, but only 5% for the whole day, bearing in mind that there are two peak times: early in the morning when the sun has not yet risen and in the evening when it has already set. The cleaning operation is very important but very difficult. Indeed, because train and network maintenance has to take place at night, the only time cleaners can intervene and panels can be installed is between 1 and 4 o’clock in the morning. This obviously represents a logistical, technical and safety challenge for the night workers. Upkeep and maintenance are subcontracted to the company (Sun Edison) which has installed the panels under a 25-year maintenance contract.


PV plant on the DMRC workshop roof

Aerial view of the Yamuna Bank depot, with its roofs covered in PV (Google)

Whilst the government incentives cited above applied here too, we could see that DMRC’s mobilisation went further and showed a veritable appropriation of the self-production objective in order to reduce operation costs, of which electricity represents a big proportion. Our interlocutor further stressed that the public too had a direct interest in this since a reduction in outgoings translated directly into a reduction in production costs and therefore in the fares thepublic was charged (or into a reduction in subsidies, one might think). He informed us that discussions had therefore been going on with the regulating authorities: DMRC would wish to break free of DISCOMs, who have local monopoly over the distribution of electricity, as it would like to be able to develop its own production capabilities outside Delhi. For example, it already has a supply agreement with a solar park in Madhya Pradesh (under the RESCO model) from which it buys electricity at a price lower than the DISCOM one (around Rs 2 difference), and this despite the charges imposed by the Central Electricity Regulatory Commission (interstate charges) and the Delhi Electricity Regulatory Commission (intrastate charges). One may think, subject to confirmation,that this evolution of things would not be favourable to DISCOMs’ business model since, thanks to these adjustments, big industrial clients such as DMRCplay an essential role in the implementation of a policy for the social integration of small consumers (for further details see Kohler, Zerah 2013). This most definitely explains why DMRC was refused exemption from the interstate andintrastate charges. The same goes for DMRC’s failed application – it intends to reapply – to obtain a “deemed licensee” status (already granted to the railways) which would allow it to develop a strategy forachieving complete autonomy.

Gurgaon: solar condos


View of a part of the PV plant on the roof of the Wellington Estate condominium (Eric Verdeil, CC)

We witnessed another very impressive example of solar equipment in a condominium of around 500 apartments in Gurgaon, a new city south of Delhi developed in thepast 20 years to serve the needs of the tertiary industry. In the past twoyears, the estate had been investing in an initial stage of PV panels representing 200 kW. The plant covered around 50% of the roofs of the residence. Thanks to the success of the operation, a new phase is being planned.

There are several favourable factors worth mentioning: first of all, the personality of the estate’s current president seemed decisive. This businessman who had retired from the advertising sector and had nothing to occupy him (his children were abroad and his wife had her own activities) wanted to find savings for the estate and do something for the environment even though he did not define himself as an activist. As it happened, the implementation of the project had led to a reduction in costs, not least because his style of management seemed more widely geared toward the rationalisation of outgoings – lighting in the communal areas had already been switched over to LED bulbs, which had resulted in a big saving comparable to the amount saved by the solar production. Further, we can again cite the regulatory incentives (net metering, etc.). Above all else, this estate, like many of its counterparts in Gurgaon, had a long history of infrastructural autonomy. It had been equipped with water wells, which now seemed to play only a supporting role, but also with generators, which at first used to supply all the electricity within the complex. If today this production capability only played a back-up role in case of power cuts (very short ones amounting to around 10 hours a month), this was the result of the relatively recent improvements in electricity supply made by DISCOM. In this context, the solar investment was part of a continuum, even if it did not in and of itself represent a new back-up capability – the solar system could work in conjunction with the public grid or the generators, but not on its own.

Aerial view of the Wellington Estate solar plant (source: YouTube, Wellington Board, uploaded on 6 December 2017)

A truly remarkable characteristic was the estate’s desire to customise this technology and guarantee the quality of the work carried out. For this purpose, and despite the opposition of many residents, the estate’s president and a couple of other individuals involved had worked together with the selected consultant to improve the project. Indeed, they had chosen a small company (Green Roots) over a bigger one that provided only standard solutions. The president actually had an interesting view on the circumstances that would allow this solar sector to take off. According to him, there should be third-party consultants who could help residences to negotiate with the companies. Indeed, he told us that their discussions with Tata or Hero Solar had failed because they had refused any design changes or had wanted to impose their own choice of panel supplier. Finally, even though the supplier was supposed to be in charge of maintenance for five years, in reality it was the estate that would take care of the upkeep and maintenance of the plant.

In economic terms, they did not hesitate to invest for the long term by buying good-quality material and now they were able to reduce their electricity bill by 10%. They planned to use their monitoring system to optimise their plants. This was part of a wider optimisation strategy (installation of more energy-efficient ventilators, motor replacement, etc.) and in-house management, which, according to the president, explained why their outgoings were far lower than those of other Gurgaon residences. This, he maintained, would explain the end of the protests witnessed at the beginning of the project.

To what extent could this situation be replicated in other parts of the city? This question seemed all the more relevant since CPR’s work on the Smart City programme shows that the development of solar roofs is a clearly stated priority for Gurgaon. Nevertheless, according to our interlocutor, the city had not provided any support and its achievements in this area seemed more or less non-existent. Only 6 or 7 other building complexes seemed to have developed similar investment programmes, but less successfully and especially with a lower degree of customisation. And yet, the lowering of installation costs seems a powerful incentive, albeit expertise is also required, something that not all estates and their subcontractors have.

The rapid electrification of Delhi’s slums and working-class neighbourhoods

Rémide Bercegol and Gowda Shankare took us to visit two slum areas, which were also waste collection and treatment hubs. First we went to Hanuman Mandir, Rama Krishna Puram, a completely illegal neighbourhood located in a low area with a nala, an open-air sewer, running across. Many of the houses there were made of scrap materials. The collected waste, which was deposited in huge enclosures or just along the sewer, was sorted and then sent to wholesalers who would then negotiate with the recycling industries. We were greeted by the periwalah, the entrepreneur in charge of this trade, who was also the sarpanch or head of the neighbourhood. Collectors deposited their bales on a scale outside his house which measured the quantities collected every day. This dwelling, made of brick and cement with steel girders, was connected to the power grid, just as the rest of the neighbourhood was, as attested by the perfectly visible meters. Furthermore, our interlocutor received the bills and was in charge of their distribution to the residents. We found it hard to understand what the amounts quoted corresponded to, some of which seemed very high to us. Could they perhaps have been arrears?

Our host also pointed out the recently installed public lighting, which was in keeping with a national policy that had been presented to us at the CPR/AFD/CERI seminar. It involved the use of LED lights which allowed work to go on well after night had fallen.

Rémi de Bercegol highlighted that the AamAadmi (Common Man) Party’s recent access to power (the State Assembly) could explain the development of these infrastructure networks in an otherwise clearly informal neighbourhood: since this party’s coming to power in Delhi in 2012, the policy of repression against slums had lost in its intensity and certain residents who no longer feared being evicted had added an upper storey to their houses.

When asked about the consistency in supply, our interlocutor estimated that power cuts were rare and that there had been clear improvement in the past few years. However, they seemed to happen almost on a daily basis, albeit for very short periods of time and they didn’t cause any inconvenience nor did they require the acquisition of complementary devices.

The second neighbourhood was in Okhla, south Delhi. We visited several recycling workshops that dealt particularly in glass and metal. The foremen there told us that their machinery was supplied with electricity and that they were hardly disrupted by the rare power cuts. They did not have any generators.

A recently installed electric meter (Eric Verdeil, CC)

Unauthorised urbanisation on agricultural land, but the buildings were connected to the grid (Eric Verdeil, CC)

A generator outside a furniture workshop (Eric Verdeil, CC)

Behind these workshops, in an area where vegetables were still being grown, a new neighbourhood had been developing along elevated streets. Small developments,workshops, sometimes individual houses had appeared visibly fast, according to Rémi de B. and Gowda S. who had regularly been there in the past year. There too we were struck by the electrification under way, as attested by theelectricity cables and meters suspended from pylons or attached to houses. We also visited a furniture workshop which was equipped with modern machinery and supplied with electricity. A brand new generator stood grandly outside theentrance. Bought only five months previously, it was, our interlocutors told us, seldom used. Nevertheless, it was interesting that they felt the need to have aback-up plan. We also noticed an electric rickshaw whose battery was being recharged.

Urbanisation and electrification moving forward in tandem (EricVerdeil, CC)

A bit further way, on a road under construction (along a natural-gas pipeline), we had a long discussion with an estate agent about the neighbourhood’s urbanisation rationale. This urbanisation was in fact illegal as the neighbourhood was not part of the master plan. However, the agricultural land had been bought by developers who planned a (very dense) housing estate and many transactions had been conducted which, according to the agent, had been marked by various abuses. As far as access to the infrastructures was concerned, he also pointed out that electricity theft happened and was tolerated (indeed, therewere some suspicious-looking connections attached to a cable just behind us) and that, whilst the electricity distribution company had equipped the neighbourhood, the same could not be said for the supply of water which depended on a public operator. Hence residents relied on drilling into the water table.

“Bengali” camp. The satellite dish no longer workedfollowing the disconnection of an illegal electrical network by thedistribution company (Eric Verdeil, CC)

We continued our field trip by going through nearby areas inhabited by migrantswho called themselves “Bengali” but who were very likely illegalimmigrants from Bangladesh. They too were involved in waste collection and sorting for the purposes of recycling. These populations’ situation was extremely precarious and their camp dwellings were equally very basic: wood, sheet metal, fabric, plastic. We noticed a communal water point (a hand pumpwhich it was unclear whether it pumped from the water table or the nearby public network of the resettlement-colony of Madanpur). We noticed a satellite dish outside a dwelling but the residents told us it no longer worked as the electric connections they had installed had been withdrawn by the distribution company after a fire had been started by a short circuit in another camp.

These visits in slums and working-class neighbourhoods showed that electrification was advancing very fast, despite the areas’ land situation. Furthermore, theimprovement in supply meant that residents seemed to rely mainly on electricity– if they could afford it – without needing other sources of electricity production.

Dharuhera: hybrid setups in the development corridor south of Delhi

Former Rajput palace in Dharuhera (Eric Verdeil, CC)

This rapidly growing town (around 35,000 inhabitants currently, compared to 18,000 in 2011) is located in a development corridor south of Delhi: it is home to industrial areas and residences meant for middle classes and workers linked to these new economic activities. Its social and physical structure is marked by the juxtaposition of these new groups with the old village. Politically, a family of landowners, which used to be linked with the British (they organised the conscription), dominates the municipal bodies. As it happened, we were received in turn by the former mayor in his villa, then by his aunt, the current mayor, in her big modern house in the village centre, close to the neglected former palace.}

The council had recently been modernised. It occupies a big modern building where we met the executive secretary who was in fact a representative of the state. There too we learnt that the electricity supply had improved considerably and that power cuts were residual. Nevertheless, a generator outside the building had to regularly be switched on. Our municipal interlocutors mentioned the development of photovoltaic panels and the incentives from or even obligation by the stateto have them installed. The council was therefore looking at the possibility of installing them on its roof and even considered having a bigger-sized plant so as to draw additional income from it. However, we noticed that within the council there was little technical and legal grasp of these issues. We were told about an owner who apparently had had panels installed in the town and about several plants in the surrounding villages. Even if there seemed to be awell-established installer in the area (whom we were unable to meet), we saw few panels on top of the buildings. Further investigations would be required.


Urban landscape in Dharuhera (E. Verdeil, CC)}


Waste ground and, in the background, urbanisation in the form of apartment blocks in Dharuhera (E. Verdeil, CC)


A first converging observation concerns the impressive improvement in the access to electricity, in keeping with the increase in electricity production and the extension of the distribution networks. The once chronic power cuts seem today to be residual. Back-up practices are still in use by a proportion of the companies and of the middle- or upper-class households (generators, inverters) but do not concern the poorest.

The second major factor is the arrival of photovoltaic plants for production at an individual or a company and condominium level, in response to state incentives and regulatory changes, but also in order to benefit from the economic advantages provided by this increasingly approachable technology. We may well wonder whether a habit of managing shortages and disrupted supply represents afactor that favours investment in these technologies.

Report written by Eric Verdeil with additional material by Bérénice Girard, Rémi de Bercegol and Marie-Hélène Zérah.Translated from the French by Monica Biberson thanks to AFD funding to Sciences Po-CERI.

Observations sur quelques hybridations électriques à Delhi et dans sa périphérie

Le projet de recherche Hybridelec a commencé à explorer le terrain indien, en particulier la ville de Delhi et sa périphérie sud, à l’occasion de visites collectives menées à la suite du séminaire co-organisé par l’AFD, le CPR et le CERI au CPR le 30 octobre dernier (voir compte rendu séparément). Ces visites de terrain (Marie-Hélène Zérah, Bérénice Girard, Rémi de Bercegol, Gowda Shankare, Federico de Lorenzo et Eric Verdeil) ont permis d’identifier cinq configurations socio-techniques et d’observer les formes de production d’électricité et d’accès à cette ressource qui les caractérisent.

Toits solaires des bâtiments publics

Installation photovoltaïque sur le toit de la NPC, Dehli (Eric Verdeil, CC)

Une première configuration concerne de nombreuses administrations et entreprises publiques qui se dotent d’équipements solaires installés sur leurs toits. Nous avons ainsi visité, à South Delhi, dans le quartier des administrations, la National Productivity Corporation (NPC), une entreprise publique qui intervient comme consultant pour aider des administrations ou des entreprises à améliorer leur productivité, entre autres sur le plan de leur politique environnementale ou énergétique. A notre arrivée, l’alimentation électrique était interrompue, une situation tout à fait inhabituelle selon nos hôtes. Néanmoins, un groupe électrogène dédié alimentait l’ascenseur et l’éclairage d’urgence.

Autres installations sur les toits de bâtiments voisins dans le Delhi administratif (Eric Verdeil, CC)

Les responsables du département Environmental management au sein de la NPC nous ont montré l’installation d’un ensemble de panneaux voltaïques sur leur toit, pour une capacité totale de 30 kW, datant de 2015. Ce système est relié au réseau haut voltage qu’il alimente. L’électricité produite n’est donc pas consommé directement dans cet immeuble mais le NPC bénéficie d’une diminution de sa facture électrique à hauteur du nombre de KW injectés dans le réseau (haute tension). L’électricité générée représente environ 10% de la consommation annuelle de l’immeuble (30.000 Rs par mois): c’est donc une économie pour cette entreprise. Nos interlocuteurs ont beaucoup insisté sur la nécessité d’un nettoyage régulier pour éliminer la poussière et maintenir une bonne capacité de production. Aux alentours, plusieurs autres immeubles sont également équipés de panneaux de même type, quoique beaucoup moins bien entretenus. Sans que cela soit très clair, nos interlocuteurs mettent en avant plusieurs facteurs incitatifs pour expliquer cet équipement récent :

  • une décision du gouvernement central tendant à obliger les administrations à faire ce type d’installation;
  • la mise en place d’une réglementation favorable aux compteurs doubles-flux (netmetering) [mais ici ce n’est pas le cas puisque l’électricité produite alimente directement le réseau HT];
  • La réalisation de ce projet a été facilitée par l’obtention d’une subvention du Ministère des Energies Renouvelables dans le cadre de la National Solar Mission.  
  • Une autre incitation pour ce projet est l’obligation d’installer des panneaux solaires pour pouvoir être considéré comme un “green building”.
  • Au total, ces incitations financières sont substantielles: dans le cas de ce projet d’un coût total de 2.157.000 Rs, la subvention reçue fut de 647.000 rs.
  • Enfin, la baisse très rapide du coût des panneaux qui a été divisé par 2 depuis 2015 est un autre facteur essentiel de l’investissement.

La politique d’installation de panneaux PV de la Delhi Metro Rail Corporation (DMRC)

La DMRC représente un deuxième exemple très impressionnant d’installation de panneaux PV. Nous visitons un dépôt-atelier situé à l’est de la rivière Yamuna , dans le quartier de Karampura. Notre interlocuteur nous montre une installation de 50 kW ?? situé sur le toit mais en réalité, l’ensemble des toitures du complexe, y compris celles de forme arrondie, donc peu propice, sont couvertes et plus largement les stations de métro, ainsi que divers terrains libres. C’est le résultat d’une série de contrats successifs depuis 2015, avec une capacité installée qui atteint en 2018 24 MW, et une nouvelle phase d’installations est en cours. L’enjeu pour la DMRC est de produire une partie du courant dont elle a besoin pour alimenter la traction et l’éclairage de ses installations. Nous apprenons que la production solaire représente 30% de la consommation diurne, mais seulement 5% sur l’ensemble de la journée, sachant que deux pics de transport se produisent au petit matin et en soirée, lorsque le soleil n’est pas levé ou est déjà couché. L’enjeu du nettoyage est très important mais très contraint. En effet, en raison de la maintenance des trains et du réseau qui est nécessairement nocturne, le seul moment où les nettoyeurs peuvent intervenir et où les installations de panneaux sont possibles se situe entre 1 et 4 heures du matin. Cela constitue évidemment un défi logistique, technique et sécuritaire pour les employés qui travaillent de nuit. L’entretien et la maintenance sont confiés à l’entreprise (Sun Edison) qui a posé les panneaux via un contrat d’entretien sur 25 ans.

Si les incitations gouvernementales évoquées ci-dessus sont aussi valables, nous comprenons aussi que la mobilisation de la DMRC va au-delà et témoigne d’une véritable appropriation de l’enjeu d’auto-production, en vue de réduire ses coûts de fonctionnement, dont l’électricité représente une part majeure. Notre interlocuteur souligne du reste qu’il y a aussi un intérêt direct du public, puisque la réduction des charges se traduit directement dans la baisse des coûts de revient et donc des tarifs offerts au public (ou dans une baisse des subventions, peut-on penser). De ce fait, il nous fait part d’un débat avec les autorités régulatrices: la DMRC souhaiterait s’affranchir des DISCOMS qui ont le monopole local de la distribution d’électricité: la DMRC voudrait pouvoir développer ses propres capacités de production en dehors de Delhi. Elle a déjà par exemple un contrat d’approvisionnement avec un parc solaire au Madhya Pradesh (via le RESCO model) auquel elle achète l’électricité à un prix inférieur à celui de la DISCOM (environ 2 Rs. d’écart), et ce malgré l’imposition de frais imposés par la Central Electricity Regulatory Commission (interstate charges) et la Delhi Regulatory Commission (intrastate charges). On peut penser, sous réserve de vérification, que cette évolution ne serait pas favorable au modèle d’affaire des DISCOMS car les gros clients industriels comme la DMRC jouent un rôle essentiel, à travers la péréquation, pour permettre de mener une politique socialement intégratrice auprès des petits consommateurs (voir pour plus de détail Kohler, Zerah 2013). C’est très certainement ce qui explique le refus opposé à la DMRC d’être exemptée des interstate et intrastate charges. De même la demande, que la DMRC compte réitérer, pour obtenir le statut de “deemed licensee” (obtenu par les chemins de fer)  et qui permettrait de construire une stratégie d’autonomisation complète, n’a pas été accordée.

Gurgaon: des condos solaires

Vue d’une partie de l’installation PV sur le toit du condominium Wellington Estate (Eric Verdeil, CC)

Nous avons observé un autre exemple très impressionnant d’équipement solaire dans un condominium d’environ 500 appartements à Gurgaon, ville nouvelle à vocation tertiaire qui s’est développée depuis 20 ans au sud de Delhi. La copropriété a investi depuis deux ans dans une première phase de panneaux PV représentant 200 kW. Cette installation occupe environ 50% des toits de cette résidence. Devant le succès de cette opération, une nouvelle phase est prévue.

Plusieurs facteurs favorables sont à citer : en premier lieu, la personnalité de l’actuel président de la copropriété semble décisive. Cet homme d’affaires retraité du secteur de la publicité, désœuvré (ses enfants sont partis à l’étranger, sa femme a ses propres activités, …) a voulu faire faire des économies à la copropriété et faire quelque chose pour l’environnement même s’il ne se définit pas comme un activiste. De fait, la mise en œuvre du projet se traduit par des baisses de charge, d’autant que sa gestion semble plus largement orientée vers une rationalisation des dépenses – l’électricité des parties communes a déjà été changée pour des lampes LED, entraînant une grande économie, comparable au montant économisé par la production solaire. Ensuite, on peut citer de nouveau les incitations réglementaires (netmetering, etc.) Surtout la copropriété, comme beaucoup de ses semblables à Gurgaon, a une histoire ancienne d’autonomie infrastructurelle. Elle était équipée de forages pour l’eau, qui aujourd’hui n’ont semble-t-il plus qu’un rôle d’appoint, mais aussi des groupes électrogènes qui à l’origine assuraient la totalité de l’alimentation électrique du complexe. Si aujourd’hui cette capacité de production n’a plus qu’un rôle de back up en cas de coupure (environ 10 h par mois, sous la forme de petites coupures très courtes), c’est le résultat de l’amélioration de fourniture électrique par la DISCOM et c’est une situation relativement récente. Dans ce contexte, l’investissement en solaire s’inscrit dans une continuité même si il ne représente pas, en lui-même une nouvelle capacité de back up. Le solaire peut fonctionner couplé avec le réseau public ou les générateurs, mais pas seul.

Vue aérienne de l’installation solaire de Wellington Estate (source: YouTube, Wellington Board, mise en ligne 2 décembre 2017)

Une caractéristique très remarquable est la volonté de la copropriété de s’approprier cette technologie et de garantir une qualité des travaux. Pour cela, et malgré une opposition de nombreux résidents, le président de cette copropriété et une ou deux autres personnes impliquées ont travaillé avec le consultant sélectionné pour améliorer le projet. Ils ont d’ailleurs préféré une petite entreprise (Green Roots) plutôt qu’une entreprise plus large qui propose des offres standardisées. Il a d’ailleurs une perspective intéressante sur les conditions qui permettraient de faire décoller ce segment du solaire. Il faudrait selon lui des tierce parties (third party consultants) qui permettraient d’aider les résidences à négocier avec les entreprises. En effet, il raconte que leurs discussions avec Tata ou Hero Solar ont échoué car ces dernières ont refusé tout changement dans le design ou voulaient imposer le fournisseur de panneaux. Enfin, même si le fournisseur est en charge, pour cinq ans, de la maintenance, en pratique c’est la copropriété qui réalise l’entretien et la maintenance des installations.

En termes économiques, ils n’ont pas hésité à investir sur la longue durée en achetant du matériel de qualité et aujourd’hui, ils arrivent à réduire leur facture d’électricité de 10%. Ils ont pour projet d’utiliser leur système de monitoring pour optimiser leurs installations. Cette action fait partie d’une stratégie d’optimisation plus large (installation de ventilateurs moins consommateurs d’énergie, remplacement des moteurs, …) et d’une gestion in-house, qui explique selon lui que les charges sont très inférieures aux autres résidences de Gurgaon. Ceci expliquerait selon lui la fin des protestations constatées au début du projet.

Dans quelle mesure cette situation se reproduit-elle ailleurs dans la ville? La question semble d’autant plus pertinente que les travaux du CPR sur le programme Smart City montrent que pour la ville de Gurgaon, le développement des toits solaire est une priorité affichée. Toutefois, selon notre interlocuteur, la ville n’a apporté aucun soutien et ses réalisations dans ce domaine semblent à peu près inexistantes. Seuls 6 ou 7 autres complexes immobiliers semblent avoir développé des programmes d’investissement similaires, et encore, avec moins de réussite et surtout moins d’appropriation. Pourtant, la baisse des coût d’installation semble un facteur incitatif puissant, même si une maîtrise est aussi nécessaire, que toutes les copropriétés et leurs sous-traitants ne maîtrisent pas.

L’électrification rapide des slums et quartiers populaires de Delhi

Emmenés par Rémi de Bercegol et Gowda Shankare, nous visitons deux quartiers de slums qui sont aussi des points centraux dans la collecte et le traitement des déchets. Le premier, à Hanuman Mandir, Rama Krishna Puram, est un squat situé dans une zone basse traversée par un nala, un égout à ciel ouvert, un quartier est tout à fait irrégulier et dont beaucoup des maisons sont faites de matériaux de récupération. Les déchets collectés, déposés dans de vastes enclos ou simplement le long de l’égout, y font l’objet d’un premier tri avant d’être envoyés auprès de grossistes qui les négocieront avec des industriels du recyclage. Nous sommes accueillis par le periwalah, l’entrepreneur qui contrôle ce commerce, et qui en même temps est le sarpanch, le chef du quartier. Devant sa maison, les récupérateurs viennent déposer leurs ballots sur la balance mesurant les quantités collectées chaque jour. Cette habitation, en brique et ciment avec poutres d’acier, est connectée au réseau électrique, comme le reste du quartier, comme en atteste les compteurs bien visibles. Du reste, notre interlocuteur reçoit les factures et est chargé de les distribuer aux habitants. Il nous est difficile de comprendre à quoi correspondent les sommes visibles, qui dans certains cas nous paraissent très élevées. S’agit il d’arriérés de paiement?

Notre hôte met aussi en avant l’éclairage public récemment installé, ce qui correspond à une politique nationale, dont nous avons eu une présentation lors du séminaire CPR-AFD-CERI. Il s’agit de lampes LED sous lesquelles le travail se poursuit alors que la nuit est déjà bien tombée.

Rémi de Bercegol souligne que le récent accès au pouvoir (à l’assemblée de l’Etat) du groupe politique Aam Aadmi Party (parti de l’Homme commun)  peut expliquer le développement de ces réseaux d’infrastructure dans un quartier par ailleurs si clairement informel :  depuis l’arrivée de ce parti au pouvoir à Delhi en 2012, la politique de répression vis-à-vis des bidonvilles s’est amenuisée et certains habitants ont ajouté un premier étage à leur habitations, ne craignant plus une éviction.

Interrogé sur la régularité de l’alimentation, notre interlocuteur considère que les coupures sont rares et qu’il y a eu une nette amélioration depuis quelques années. Elles semblent néanmoins quasi-quotidiennes, mais de très courtes durées et ne suscitent pas de gêne ni n’obligent à l’acquisition de dispositifs complémentaires.

Le second quartier visité se situe à Okhla, au Sud de Delhi. Nous visitons plusieurs ateliers de recyclage, notamment de verre et de métal. Les contremaîtres sur place nous indiquent que leurs machines sont alimentées par l’électricité, et qu’ils ne sont guère perturbés par les rares coupures. Ils ne disposent pas de générateurs.

Un générateur devant un atelier de meubles (Eric Verdeil, CC)

A l’arrière de ces ateliers, dans un espace encore maraîcher, se développe un nouveau quartier le long de rues surélevées. Des petits immeubles de promotion, des ateliers, parfois des maisons individuelles se développent visiblement très rapidement d’après le témoignage de Rémi de B. et de Gowda S. qui passent régulièrement par là depuis un an. Nous sommes frappés, là aussi, par l’électrification en cours du quartier, visible à travers les fils du réseau de distribution et les compteurs suspendus aux pylônes ou accrochés aux maisons. Nous visitons aussi un atelier de meubles, équipés de machines modernes et alimenté à l’électricité. Un générateur tout neuf trône devant l’entrée. Acheté il y a seulement cinq mois, il sert peu selon nos interlocuteurs mais il est néanmoins intéressant qu’ils aient éprouvé le besoin de s’assurer un back up. Nous observons aussi un rickshaw à batterie électrique en train d’être rechargé.

L’urbanisation et l’électrification avancent de concert (Eric Verdeil, CC)

Un peu plus loin, le long d’une rue en construction (le long d’une canalisation de gaz naturel), nous discutons longuement de la logique d’urbanisation du quartier avec un agent immobilier. L’urbanisation est en fait illégale, le quartier étant en dehors du master plan. Mais les terrains agricoles ont été acheté par des développeurs, qui ont fait un plan de lotissement (très dense) et de nombreuses transactions ont lieu, marquées selon lui par divers abus. Concernant l’accès aux infrastructures, il souligne aussi l’existence et donc la tolérance du piratage (juste derrière nous, des connexions suspectes semblent en effet accrochées à un fil) et le fait que si la compagnie de distribution électrique a équipé le quartier, ce n’est pas le cas pour l’eau, qui dépend d’un opérateur public. De ce fait, les habitants se reposent sur des forages dans la nappe.

Campement ”bengali”. L’antenne parabolique ne fonctionne plus, suite à la déconnexion d’un raccordement électrique illégal par la compagnie de distribution (Eric Verdeil, CC)

Nous poursuivons notre visite de terrain en traversant, non loin de là, des zones habitées par des migrants qui se disent “Bengalis” mais qui sont très probablement des migrants du Bangladesh en situation illégale. Ils sont également impliqués dans la collecte des déchets et le tri en vue du recyclage. Il s’agit de populations extrêmement précaires, et leur habitat, de type campement, est également très frustre: bois, tôle, tissus, plastiques. Nous observons un point d’eau collectif (une pompe manuelle, sans savoir si elle pompe dans la nappe ou sur le réseau public de la cité de relogement de Madanpur qui passe à proximité). Devant une habitation, nous notons la présence d’une antenne satellite mais les habitants nous disent qu’elles ne fonctionnement plus, les connexions électriques qu’ils avaient installé ayant été retirées par la compagnie de distribution après qu’un incendie s’est déclenché dans un autre campement à la suite d’un court circuit.

Ces visites de slums et de quartiers populaires montrent une électrification en marche très rapide, en dépit de la situation foncière de ces zones. De plus, l’amélioration de l’alimentation fait que les habitants semblent principalement se reposer sur l’électricité – s’ils peuvent se la payer – sans avoir besoin d’autres sources de production électrique.

Dharuhera: quelles configurations hybrides dans le corridor de développement au sud de Delhi?

L’ancien palais des rajputs à Dharuhera (Eric Verdeil, CC)

Cette petite ville en forte croissance (environ 35000 habitants actuellement, contre 18000 en 2011) se situe dans un corridor de développement vers le sud de Delhi : elle accueille des zones industrielles et des résidences destinées aux classes moyennes et aux ouvriers liés à ces nouvelles activités économiques. La structure sociale et spatiale est marquée par la juxtaposition de ces nouveaux groupes avec l’ancien village. Politiquement, une famille de propriétaires terriens, qui était aussi liée aux Anglais (ils organisaient la conscription), domine les instances municipales. Nous sommes d’ailleurs successivement reçu par l’ancien maire dans sa villa puis par sa tante, maire actuelle, dans sa grande maison moderne au coeur du village à proximité de l’ancien palais en déshérence.

La municipalité a récemment été modernisée. Elle occupe un grand bâtiment moderne où nous rencontrons le secrétaire exécutif, qui est en fait le représentant de l’Etat. Nous apprenons que là aussi, les conditions d’alimentation électrique se sont fortement améliorées, les coupures étant résiduelles. Un générateur devant le bâtiment se met néanmoins en marche à intervalle régulier. Nos interlocuteurs de la municipalité mentionne le développement des panneaux photovoltaïques, et les incitations voire obligations de l’Etat à en installer. La municipalité étudie ainsi le projet d’en installer sur son toit, et envisage même de surdimensionner l’installation afin d’en tirer un revenu complémentaire. Nous comprenons toutefois la faible maîtrise technique et juridique de ces questions au sein de la municipalité. On nous parle d’un propriétaire qui aurait installé des panneaux dans la ville et de quelques installations dans les villages alentours. Même si un installateur semble avoir pignon sur rue (nous n’avons pu le rencontrer), on distingue peu de panneaux sur les immeubles. Des investigations complémentaires seront nécessaires.


Une première observation convergente a trait à l’amélioration impressionnante de l’accès à l’électricité, en lien avec l’augmentation de la production électrique et l’extension des réseaux de distribution. Les coupures de courant, naguère chroniques, semblent aujourd’hui résiduelles. Les pratiques de back up demeurent pour une partie des entreprises et des ménages de classe moyenne ou supérieure (générateurs, inverteurs) mais ne concernent pas les plus pauvres.

Le deuxième fait majeur est le décollage d’installations photovoltaïques de production individuelles ou à l’échelle d’entreprises et de condominiums, en réponse aux incitations étatiques et à la transformation du cadre réglementaire, mais aussi pour bénéficier des avantages économiques offertes par cette technologie de plus en plus abordable. On peut se demander si l’habitude de gérer des pénuries ou une alimentation perturbée constitue un facteur favorisant l’investissement dans ces technologies.

Compte rendu rédigé par Eric Verdeil et complété par Bérénice Girard, Rémi de Bercegol et Marie-Hélène Zérah.