As EV adoption accelerates worldwide, developing strategic and well-planned EVCI becomes a critical focus for policymakers. Ensuring equitable access to charging stations, minimising range anxiety, and optimising energy distribution are key challenges that must be addressed. Multiple nations and studies have highlighted the guidelines and deployment methods of charging infrastructure. Some of the studies are discussed in this section.

EVCI policy-planning metrics

Contemporary studies highlight different metrics that guide the deployment of charging infrastructure (Raghavan 2024). Figure 1 depicts the most prominently adopted metrics, which include:

• Vehicle ratio-based – 1 EVCI for every 10 EVs (IEA 2024)
• Distance-based – ensuring stations are placed at specific intervals (50, 100 km, e.g.) (Gnann et al. 2018, Shoman et al. 2023)
• Location-based energy allocation – matching infrastructure capacity with EV demand and power needs by charging location (depot, on-route or highway corridors, for example) (Trafikverket 2019, 2021a)
• Density or coverage-based – ensuring adequate chargers per road network or population in high-utilisation areas.

Table 2. Snapshot of global Electric Vehicle Charging Infrastructure (EVCI) roll-out targets and provisions in major EV markets as of April 2024

Source: EU 2022, EC 2023; Government 2023; (T&E 2021, BMDV 2022; US DOT 2021; MoP 2022, BEE 2023, 2024;

Bloomberg 2023, NEDO 2021

Note: PCS/P: public charging stations/points or ports; HRS: hydrogen refuelling station; TEN-T: Trans-European transport network. Targets and provisions summarised in this table are non-exhaustive.

Globally, advanced EV markets have adopted different policy frameworks to accelerate EV infrastructure deployment, with the United States and the European Union serving as key case studies. The US National Electric Vehicle Infrastructure (NEVI) programme follows a corridorbased approach, ensuring fast chargers are placed every 50 miles along major highways, particularly focusing on enabling long-distance travel and freight movement (Stenstadvolden 2024). Meanwhile, the EU Alternative Fuels Infrastructure Regulation (AFIR) enforces a mandate-driven, command-and-control approach, requiring charging stations at least every 60 km along the Trans-European Transport Network (TEN-T), along with interoperability and standardisation measures. These global strategies offer valuable insights for India, where a hybrid approach balancing corridor-based development with regulatory mandates could accelerate EVCI deployment, ensuring efficient, scalable, and sustainable electrification of national highways and urban networks.

The US NEVI, EU AFIR, and India’s MoP guidelines

The National Electric Vehicle Infrastructure (NEVI) programme, launched under the Bipartisan Infrastructure Law (BIL), is a cornerstone of the US government’s strategy to accelerate EV adoption. The programme follows a corridor-style planning approach, with a target of deploying charging infrastructure across 50,000 miles (graphic says 75,000 miles) of designated corridors, to ensure EV drivers have access to reliable charging stations along major highways and intercity routes. A key feature of NEVI is its minimum coverage requirement, which mandates that fast-charging stations be installed every 50 miles (80 km) along designated alternative fuel corridors (AFCs) and within one mile (1.6 km) of highway exits (J. Chu, Gilmore, B., Hassol, J., Jenn, A., Lommele, S., Myers, L., ... & Shah, M. 2023). These stations must offer at least four DC fast chargers (minimum 150 kW each) to support high-speed charging for passenger vehicles and commercial fleets. The approach prioritises range confidence and seamless travel, addressing one of the biggest barriers to long-distance EV adoption.

Funding under NEVI is structured to incentivise state governments, private players, and utilities to invest in EVCI. States are required to submit detailed EV infrastructure deployment plans, ensuring that charging networks align with expected EV growth. Additionally, NEVI promote public-private partnerships (PPPs) to encourage sustainable business models, aiming to reduce dependence on government subsidies over time. The programme also aims to integrate renewable energy sources and grid-resilience measures, enhancing the system-wide sustainability (US DOT 2021).

The European Union (EU) Alternative Fuels Infrastructure Regulation (AFIR) represents a regulatory-driven approach to EV infrastructure development (EC 2023). The EU’s approach is mandate-driven, seeking to ensure that all member states comply with standardised charging infrastructure targets. AFIR sets legally binding requirements, making it a commandand-control policy framework to create a seamless and interoperable EV charging network across Europe. AFIR mandates that public charging stations be available at least every 60 km (37 miles) along the Trans-European Transport Network (TEN-T) corridors. This ensures that passenger and freight EVs can travel without range concerns. It also requires that fast chargers (minimum 150 kW) be installed for light-duty vehicles, and 350-kW chargers for heavy-duty trucks. Unlike voluntary or market-driven initiatives, this regulation compels governments and businesses to meet infrastructure deployment targets, backed by stringent reporting and compliance mechanisms (EC 2023).

A critical component of AFIR is interoperability and payment standardisation. The policy mandates that charging networks across EU member states be harmonised, allowing users to access any public charging station without needing multiple subscriptions or proprietary payment systems. This is achieved through universal charging protocols and transparent pricing mechanisms, reducing friction for EV users. Additionally, AFIR integrates hydrogen refuelling infrastructure alongside EV chargers, recognising the role of multiple alternative fuels in decarbonising transport. The regulation ensures that the entire mobility ecosystem, including long-haul freight, public transit, and private vehicles, can transition seamlessly to cleaner energy sources.

Ministry of Power (MoP) Guidelines on Charging Infrastructure comprise comprehensive standards for EV charging infrastructure deployment (MoP 2022), to ensure uniformity, accessibility, and ease of implementation. Key provisions include:

• Charging station coverage: At least one charging station every 3x3 km grid in urban areas, and every 25 km along national highways for light vehicles, and at 50- km intervals for heavy-duty vehicles
• Tariff regulations: Charging point operators are treated as de-licensed entities, allowing market-driven tariff structures while ensuring grid stability
• Public-private participation: Encourages PPP models and state-led initiatives to accelerate infrastructure rollout
• Battery swapping: Recognised as an alternative to fixed charging solutions, particularly for fleet and last-mile delivery segments
• The PM-EDRIVE initiative (Ministry of Heavy Industries 2024) has set an aggregate target of 1,800 fast chargers rated 240 kW along 20 prominent truck corridors and 35 national highways with heavy car and bus traffic

Studies on the electrification of heavy-duty vehicles in the US (Borlaug 2022, NACFE 2019) and the EU (Gnann et al. 2018, Shoman et al. 2023, Speth and Plötz 2024) indicate that depots will play a similar role as home charging for passenger cars. Overnight depot chargers rated up to 50 kW are expected to be the major charging location, fulfilling 60–80 per cent of the charging requirements. High-power chargers (HPC) up to 150 kW and ultra-fast chargers (UFC) rated up to 350 kW are estimated to meet 15–30 per cent and 5–10 per cent of the total charging demand, respectively (Trafikverket 2021b; Mathieu et al. 2020; Noordjik, Refa, and Rookhuijzen 2020). Studies suggest a 1:1 ratio of Battery electric trucks (BETs) to overnight depot chargers (50 kW), 1–4 high power chargers (HPC up to 150 kW), and 10 ultra-fast chargers (UFC up to 350 kW) for heavy-duty (HD) truck fleets (Noordjik, Refa, and Rookhuijzen 2020; Ricardo 2020; Speth and Plötz 2024)

Both NEVI and AFIR offer valuable lessons forIndia’s EV infrastructure planning. While NEVI’s corridor-style approach is well-suited to India’s Golden Quadrilateral and major freight routes, AFIR’s mandate-driven structure could provide a policy blueprint for standardising and accelerating EVCI deployment nationwide. Integrating insights from these global models will be key to developing a robust, scalable, and future-proof EV charging network in India. India has introduced several policies and schemes to accelerate the adoption of electric vehicles (EVs) and strengthen the Electric Vehicle Charging Infrastructure (EVCI). Recognising that infrastructure readiness is a key enabler for EV transition, the government has implemented financial incentives, regulatory frameworks, and operational guidelines to drive EV adoption across various vehicle segments. India’s multi-pronged approach to EV adoption and infrastructure development through FAME, PM-EDRIVE, and MoP guidelines demonstrates a strong commitment to decarbonising transport and reducing dependence on fossil fuels. However, ensuring efficient execution, addressing infrastructure bottlenecks, and integrating renewable energy into charging networks whilst considering their financial viability will be key.

EVCI sizing and siting criteria

Multiple studies have examined factors considered for the siting of EV charging stations. These include economic and environmental parameters, besides those concerning transportation, energy, urbanisation, existing resource utilisation, and driver convenience. The economic factors include capital, maintenance, and operational costs. Environmental parameters include green areas, slopes, surface temperatures, with vulnerability to natural disasters like floods, landslides, and earthquakes also considered. The transportation-related factors include vehicular flow, junctions, parking and transit zones; and energy-related parameters pertain to electricity grids and substations, and the renewable energy potential of existing land parcels. The urbanisation factors include the proximity/availability of rest stops, industrial clusters, commercial zones, fuel stations, and other amenities. Existing resource utilisation means the current utilisation rate of existing chargers. As this dataset is very critical to get due to the nascent age of the industry, we are not considering it for this study. User convenience includes user accessibility, wait time, and comfort during charging halts. Table 3 highlights all the factors considered in contemporary studies.

The most effective spatial analysis methodologies employ multi-criteria decision analysis (MCDA) frameworks that evaluate potential charging locations across numerous dimensions. These approaches typically integrate geographic information systems (GIS) with weighted overlay techniques to generate comprehensive suitability maps. Studies on improving EV charging point placement show that these techniques use ‘boolean mask layers’ to remove unsuitable locations, and weighted suitability factors to pinpoint ideal sites (Mahdy 2022). Advanced geo-statistical frameworks incorporate both exclusionary constraints and positive suitability factors. For instance, a study seeking to develop a ‘walkable factor grid’ analysed 27 distinct input factors, including land use, demographics, and employment centres, to generate hot spot maps that visualised high, moderate, and lowsuitability areas for charging infrastructure (Zhang 2018). For high-power charging along transportation corridors, specialised analysis techniques have been developed that focus on strategic placement to enable long-distance travel. The EV Fast-Charging Corridor Road Map in the US identified corridor gaps by analysing driving distances between charging stations, categorising segments as fully compliant (less than 25 miles to charging), partially compliant (25-50 miles), or non-compliant (more than 50 miles) (Carbon Solutions & Great Plains Institute 2023).

Table 3. Economic factors primary concern in EVCI sizing-and-siting research

Study objectives

This study seeks to address four research questions:

• Which MHDTs are best suited as the first 100 pilot candidates for ZET transition along the Delhi–Agra corridor, and how to screen them based on their characteristics, use-case and operational profiles?
• Where should EVCI be deployed along NH44 to maximise suitability, and how can locational intelligence inform these decisions?
• In a scaled-up 30 per cent ZET scenario, what are the comparative sizing, siting, economic and environmental implications of two distinct EVCI build-out strategies — anchor node or ‘hubs’, and distributed network or ‘corridor-wide’?
• Does the current ZET product and market landscape meet the expectations of LSPs? If not, what are the major adoption barriers, enablers and policy expectations?

The rest of the report is organised as follows: Section 3 describes the data, analytical framework and methodology. Section 4 summarises key findings from fleet-wide operational insights to the comparative assessment of the two EVCI build-out strategies. Section 5 lists near-term priority interventions necessary to enable ZET uptake. Using the insights from this study, Section 7 provides policy recommendations to support long-term ZET transition.

Overview of the Delhi-Agra corridor (NH44)

The Delhi-Agra corridor (NH44) is designated by the Ministry of Power (MoP) and Ministry of Heavy Industries (MHI) as one of the priority traffic corridors (truck, car, and bus). This NH corridor spans two major interchanges—Badarpur and Palwal—and five toll plazas, interlinking industrial clusters, logistics parks, and two major tourist hubs in Mathura/ Vrindavan and Agra.

The Delhi-Agra highway is a part of the Golden Quadrilateral segment that connects Delhi and Kolkata. It also provides connectivity to the east-west and north-south corridors of the national highway development plans (India). It spans approximately ~250 km, stretching from the Badarpur elevated toll plaza in Delhi to the NH19 Guru Ka Taal in Agra. This highway traverses three states—Delhi, Haryana and Uttar Pradesh—and passes through several cities and towns, including Faridabad, Palwal, Hodal, Kosi Kalan, Mathura and Agra. There are five toll plazas along the route: the Badarpur elevated toll plaza and Badarpur toll plaza at the Delhi-Haryana border which are very nearby, the Gadpuri toll plaza between Faridabad and Palwal, the Karman toll plaza at the Haryana-Uttar Pradesh border near Hodal and Kosi Kalan, and the Mahuvan toll plaza located between Mathura and Agra in Uttar Pradesh. The highway serves as a critical connection between Delhi and Agra, apart from the Yamuna Expressway. The corridor experiences a diverse mix of traffic, which includes cars, trucks, LCVs, and buses.

Electrification potential

Doubling the range of current ZET models to 400 km would readily meet 75 per cent of MDT and 60 per cent of HDT daily driving requirements.

Figure 7 depicts the cumulative distribution function (CDF) plot of daily driving distances. Nearly 50 per cent of MDTs and 40 per cent of HDTs drive 250 km or less. However, the steep slope of both curves shows a high share of truck operations in the 250 to 400 km daily distances. Consequently, the electrification potential is highly sensitive or elastic to modest improvements in battery range. A 400 km range could accommodate the daily travel demand of over 70 per cent of MDTs and 60 per cent of HDTs. The median absolute deviation—MAD (119/142 km for MDT/HDT)—and interquartile range—IQR (276/305 km for MDT/HDT)— values indicate greater trip distance variability among HDTs. Furthermore, a notable share (25 per cent) of HDTs travel 500 km and beyond, as seen in the distribution.

Figure 7. Cumulative distribution function of daily distance driven

Source: Authors’ analysis

Note: Dashed and dotted lines show the different ZET ranges. 

Interestingly, the MAD and IQR of the average rest time per stop of both MDTs and HDTs were 15 minutes, despite HDTs making more stops (2.1 versus 1.7 for MDTs) and taking longer breaks per stop (44 versus 38 minutes for MDTs) on average. The existing dwelling patterns and stoppage frequency already offer favourable charging windows without needing major operational or intra-day schedule changes.

Key driving and dwelling parameters of MDTs and HDTs were analysed using non-parametric tests to determine if ownership type (3PL/LSPs, owned & operated or O&O, hired or leased) is a factor. For O&O fleets, differences in trip distance, number of rest stops, rest duration, and tolls were not statistically significant between MDTs and HDTs (Table A3). The contrast was observed across all the above parameters between MDTs and HDTs operated by 3PL/LSPs, perhaps alluding that a differentiated (by ownership) yet targeted (by ZET segment) strategy for electrification could be necessary.

Freight flow dynamics

The majority of trucking activity on NH44 is within Haryana and Uttar Pradesh, concentrated around the main corridor and key arterial distribution points.

The origin-destination (O-D) pattern analysis of the freight dataset obtained from the VIUS reveals that trucks move across 15 states, running from Jammu and Kashmir in the north to Rajasthan in the west, West Bengal in the east and Telangana in the south. Delhi, Haryana, Madhya Pradesh, Punjab, Rajasthan, and Uttar Pradesh account for 98 per cent of recorded NH44 trucking activity origins and destinations, establishing them as the primary hubs. Approximately 40 per cent of the freight activity occurred to and from Haryana and Uttar Pradesh. Twelve per cent of the trucks operate within Uttar Pradesh, and 10 per cent ply from Uttar Pradesh to Delhi. In Uttar Pradesh, Mathura and Agra districts emerge as key nodes,

Figure 8. 96 % of truck flow is concentrated in Delhi, and cities in Haryana and Uttar Pradesh

which lie directly along the corridor. Similarly, the Faridabad and Palwal districts play a crucial role in Haryana’s freight network. Smaller towns on the highway network, such as Hodal and Kosi Kalan, are emergent freight centres on the highway corridor. Figure 8 shows the origin and destination locations.

On average, the geodesic trip distance was 310 km (self-reported distance was 327 km, Figure A8), while the length of the corridor is approximately 250 km. This indicates that the trucks are utilising the highway corridor as well as travelling to nearby areas beyond their immediate stretch. Key nearby freight centres include Bharatpur in Rajasthan, Kanpur, Aligarh, Etawah and Firozabad in Uttar Pradesh, and Gurugram and Rohtak in Haryana. Key commodities transported along the corridor include perishables such as food grains, fruits, vegetables, and dairy products, along with cement and construction materials, making it a vital link for regional and interstate trade (Figure A5 and Figure A8). These observations indicate intense trucking activity through the corridor and the peri-urban area. The spatial clustering of truck O-D pairs and their adjoining districts confirms high route predictability and a concentrated tonne-kilometre share on NH44.

The first 100 ZET candidates

Convergence of haulage, commodity demand and operational fit offers the highest entry point for early ZET adoption.

Figure 9 shows the use-case, axle configuration, and tonnage profile of the first 100 highpotential ZET candidates identified according to the chosen screening criteria. Close to 60 per cent of these are MDTs, of which 2-axle MDTs with daily driving distances in the 200–300- km range account for the largest share (~35 per cent). Nearly a third are 4–6-axle HDTs operating in the 200–300-km range. Figure 10 offers another perspective on the first 100 ZETs based on the commodity transported. Construction materials and cement represent

Figure 9. ‘The First 100 ZETs’ by use-case, tonnage and MHDT segment

Figure 10. The First 100 ZETs by major commodity category transported

the largest transition opportunity, characterised by predictable routes between production facilities and major urban centres. These materials’ high density and heavy loads explain the predominance of HDTs. Food grains and pulses present the second-largest opportunity, with a notably higher proportion of MDTs compared to other categories. This reflects the regional nature of grain distribution networks, where farm-to-processor and processor-to-market routes are routine. The use case of MDTs in dairy products highlights the time-sensitive nature of this supply chain, with high-frequency, fixed-route operations between collection centres, processing facilities, and urban markets. Similarly, fruits and vegetables show balanced MDT/ HDT distribution, reflecting local and regional distribution patterns. A slightly higher share of HDTs transporting petroleum products and chemicals may be tankers plying from refineries to regional demand centres.

EVCI suitability map

High suitability zone clusters near urban and peri-urban nodes, interspersed with long stretches of moderate to very low suitability towards Agra.

The resultant composite layer reveals the integrated suitability pattern along the Delhi-Agra corridor, with distinguishable variations within specific segments between Delhi, the toll plazas at Badarpur, Gadpuri, Karman, and Mahuvan, and Agra. The spatial distribution of suitability scores reflects the combined influence of all 10 criteria—depicted in Figure 4 and described in depth in the Annexure—at each location. The resultant composite layer was classified into five distinct suitability classes, ranging from least suitable to most suitable. This classification, with its granular location data, helps stakeholders prioritise locations for infrastructure development.

• Threshold definition: Suitability classes were determined based on natural breaks and the statistical distribution of composite scores.
• Visualisation and mapping: The classified regions were mapped to provide a spatial representation of suitability, aiding in decision-making for infrastructure planning.

The five classes range from low- to high-suitability zones based on 10 factors mentioned in Table 5. The objective is to identify high-suitability clusters within low-suitability segments to highlight areas where land is affordable, demand is rising, and supporting infrastructure is present, ensuring optimal development opportunities.

Table 5. EVCI site suitability classification

Source: Authors' analysis 

The corridor, including a 500-m buffer on both sides, covers a total area of 204 sq km. Approximately 50 per cent of the corridor is not suitable for EV charging infrastructure due to the dominance of agricultural or vacant land, which lacks essential amenities, accessibility, and commercial activities Only 12 per cent can be classed as having very high suitability, and it is concentrated mainly in the Delhi to Gadpuri stretch. The Gadpuri to Agra stretch primarily features low- and moderate-suitability sites for potential EV charging locations. The overall volume of toll-plaza transactions (including cars/jeeps/vans, LCVs, and buses) decreases from Delhi to Agra (Annexure). However, daily truck volumes (2, 3, and 4-6 axles) rises sharply; near Mahuvan (~6,700), they increase five times relative to Badarpur (~1400), and, near Agra (~3300), they more than double.

Figure 11. EVCI suitability map showing high suitability clusters accompanied by stretches of low suitability

Source: Authors' analysis 

It is interesting to observe that the same stretches of highway fall under the moderateto very-low-suitability classifications. The increase in truck traffic along segments with moderate to low suitability, coupled with the presence of high freight activity (refer to O-D pairs, Figure 8), clearly underscores the need for a balanced evaluation of anchor node or hub-based and distributed network or corridor-wide EVCI build-out. Unless otherwise explicitly mentioned, the preceding analysis and observations mainly pertain to the 30 per cent ZET scenario (N=1,000 ZETs).

Existing EVCI installations

Capacity, reliability, compatibility and equitable coverage pose challenges and opportunities for kick-starting high-power and fast charging for trucks.

Figure 12. Existing EVCI along the corridor by charging kW levels

Source: https://www.statiq.in/ev-charging-station and Google Places API 

The NH44 corridor between Delhi and Agra currently hosts 16 fast-charging stations with 44 charging guns (or charge points), of which only 30 are operational, translating to a working availability of ~70 per cent (Figure 12). The cumulative installed charging capacity across the corridor is approximately 2 MW, with only 10 chargers rated 120 kW or higher. The average daily throughput, depending on the stretch, utilisation, charging power, and number of existing EVCIs on NH44, varies between 3.5 MWh and 10.5 MWh (details refer to Annexure)

However, the spatial distribution of chargers is highly uneven. The Delhi–Badarpur stretch alone accounts for six chargers, all within a zone with moderate to very high suitability. This clustering continues, albeit moderately, through the Badarpur<>Gadpuri (6 chargers) stretch, gradually dwindling over Gadpuri<>Karman (3 chargers), and appearing again in Karman<>Mahuvan (5 chargers) stretches.

This imbalance is a symptom of the broader issue of inconsistent coverage, where fast chargers are concentrated in and near urban fringes, but taper off in peri-urban and rural stretches. Considering the opportunity cost, both geographic parity and higher installed capacity are vital design aspects, especially in stretches with increasing truck throughput and suitability potential, be it the pilot or scale-up phase of ZET adoption.

Charging Demand and Supply Gap

Directional build-up of traffic and thereby charging demand is observed in areas with sparse EVCI coverage and/or site suitability constraints

Figure 13. 30% ZET: (a) truck flows and (b) average daily charging demand by NH44 stretch and ZET segment

Truck movement and estimated charging demand across the NH44 corridor segments highlight a clear concentration of freight activity between Karman and Mahuvan, whereas the existing EVCI is more concentrated in the Delhi-Badarpur stretch (Figure 11). The average daily demand was estimated to be ~370 MWh, of which the Karman-Mahuvan stretch accounted for ~35 per cent or 130 MWh (Figure 13). The demand builds progressively and consistently at every stretch from north to south, peaking at Karman–Mahuvan. Gadpuri– Karman emerges as the second-highest demand stretch with ~95 MWh, followed by Mahuvan–Agra (63 MWh) and Badarpur–Gadpuri (53 MWh). Notably, there are no chargers (120 kW and above) in the Mahuvan–Agra stretch (Figure 12).

Though MDTs account for 45 per cent of flows, they represent only 13 per cent of the overall demand. Of the 454 MDTs, 242 are expected to charge to only 50 per cent SoC, whereas HDTs exhibit a higher share of deeper charging cycles. The data shows that nearly a third (175 out of 546) HDTs charged to 90 per cent SoC (compared to 144 MDTs), while 303 HDTs charged 50 per cent SoC (compared to 242 MDTs). Given that the analysis relies on a stated response to refuelling their existing ICE variant truck as a proxy for charging, such assumptions are practically plausible, as it constitutes one among the many possible real-world choices. For instance, it is reasonable to assume that, for HDTs, any stop they make needs to be worthwhile, given that the longer the average stoppage time, the higher the opportunity cost of idling. This makes deep charging (e.g., ≥75–90 per cent SoC) rational and apropos.

MDTs are often deployed for high-frequency, low-distance runs with relatively tight delivery schedules, such as consumption goods, or shuttle logistics between nearby industrial nodes and warehousing clusters (Palwal, Faridabad, and mandis on the outer periphery of Agra, for example). In such contexts, drivers may opt to “top up” rather than refuel fully, especially if they anticipate a short turnaround or the next trip may not require a full tank. If these are assumed to mirror ZET charging practices, a 50 per cent SoC is reasonable. Hence, the prevalence of mid-tank to full-tank or, equivalently, 50 per cent SoC among MDTs reflects a practical balance between duty cycle, predictability and recurring nature of trips, and usecase alignment—where a full tank is not always necessary.

To estimate the supply gap, peak hourly demand is compared against the capacity of existing EVCI installations. This intermediate step serves as a necessary precursor to identifying where retrofitting, reinforcement, or augmentation of existing EVCI, either as stand-alone chargingonly sites or where co-located with fuel stations, without requiring major utility grid overhauls at 33 kV and above levels. Isolating and determining the supply gap that cannot be met by existing and retrofitted sites forms the basis for examining the anchor node and corridor-wide distributed network approach for EVCI build-out.

Figure 14. Peak hourly demand across the NH44 stretch-wise and total corridor-wide for various PHF

Source: Authors’ analysis
Note: Strictly speaking, from a classical travel demand and highway capacity planning perspective, PHF refers to the ratio of average volume during the peak hour period to four times average volume during the peak hour period. For simplicity and convenience, a relaxed definition of PHF is adopted, expressed as the ratio of peak hour to average daily traffic.

Figure 14 presents the hourly peak demand across the five stretches of NH44 for PHF varying from 2.5 to 25 per cent. Even at relatively low to modest PHF (5-10 per cent), corridor-wide demand reaches 30-45 MWh, indicating substantial baseline power requirements. The demand profile is not evenly distributed, with certain stretches (particularly Karman<>Mahuvan and Mahuvan<>Agra) having a noticeably higher peak demand. At higher PHF values (20-25 per cent), the overall corridor demand increases dramatically, suggesting potential grid stress points during the peak traffic period.

Figure 15. Existing EVCI capacity—supply and demand gap under select PHFs

Source: Authors’ analysis
Note: Zero values (or supply adequate) are excluded from the Y-axis in log-scale. LB/UB– lower and upper bound and charging relation assumptions are detailed in Annexure, Figure A22. The corridor peak demand assumes the worstcase coincident peak demand across the individual stretches happening at the same instance. 

Figure 15 depicts the supply gap, expressed as the difference between peak hourly demand and existing EVCI capacity across the five NH44 stretches, as well as the total corridor, under 5 per cent and 25 per cent PHF. While multiple intermediate PHF values were considered, only these two cases were elaborated to streamline interpretation and serve as exemplar bookends for planning and stress testing. At 5 per cent PHF, the supply gap is minimal to moderate and largely concentrated in the southern stretches. While segments such as Delhi– Badarpur and Badarpur–Gadpuri theoretically show no observable gap, practically ~300 kWh—800 kWh and 1 MWh–2 MWh is still required in the stretches. The total supply gap across the corridor varies from ~3 MWh to 13 MWh. In the case of 25 per cent PHF, a notable supply gap can be evidenced across all stretches, with the corridor’s demand or supply gap rising sharply to anywhere between ~50 MWh and 80 MWh. It is more pronounced in the Gadpuri<>Karman (14 MWh–21 MWh) and Karman<>Mahuvan (~17 MWh–31 MWh) stretches. The Mahuvan<>Agra stretch is completely unserved by the current EVCI installations, effectively a ‘charging desert’ or ‘blindspot’, where the entire demand defaults to a supply gap.

At higher PHFs, the demand burden shifts and spreads north: Karman<>Mahuvan and Gadpuri<>Karman together contribute over 60 per cent of the corridor’s total supply gap. Delhi<>Badarpur consistently shows low to no supply gap (3 per cent at most) even at 25 per cent PHF. It must be clarified that this seemingly marginal share could be misleading by sheer virtue of interpretation as a percentage of the corridor’s total gap. However, in absolute terms, this still translates to 2 MWh. The Mahuvan<>Agra stretch occupies a persistent and significant share (20 per cent –30 per cent) of the corridor’s total supply gap at 25 per cent PHF. The extremities in PHF were chosen to show how and where the supply gap emerges, as well as intensities and shifts within the stretches and across the entire corridor. In the subsequent analysis, a PHF of 5 per cent is chosen.

Sizing EVCI for 30 per cent ZET

Speed and future-proofing versus scale and flexibility govern the decision-making contours of EVCI build-out for ZET adoption, viz-à-viz anchor node ‘hubs’ or distributed network ‘corridor-wide’ strategy.

This section estimates the scale of infrastructure needed to support a 30 per cent zeroemission truck (ZET) uptake across NH44. Indicative and recommended sizing benchmarks are the focus over exact and precise prognosis. Due to the interrelationships between these technical aspects, siting and, consequently, costs, the two strategies are examined separately. A comparative summary of the key characteristics of the nodal and network strategy for EVCI build-out is presented in the Annexure, Table A12.

An anchor node or charging hub prioritises concentrating EVCI in fewer, high-capacity sites rated up to 6 MVA at approximately 40–50 km intervals. A distributed network or corridorwide strategy emphasises coverage using charging stations rated up to 1.5 MVA across the corridor at 10–20 km intervals. Or the chosen PHF, the peak hour demand varies between 1.3 MWh in the Delhi<>Badarpur stretch to 6.5 MWh in the Karman<>Mahuvan stretch.

Table 6. Exemplar unit specifications of nodal and network site

Source: Authors’ interpretation from (K.-C. Chu, Miller, K. G., Schroeder, A., Gilde, A., & Laughlin, M. 2024; D2Z 2023; Green 2020; Ladeinfrastruktur 2022; Unterluggauer 2022; ABB 2021; Burges 2021)
Notes: Local reinforcement refers to upgrading low-voltage (LVDS) to high-voltage distribution system (HVDS); the primary substation is 33/11 kV. Details in Annexure. 

The worst-case coincident peak demand is 18 MWh across the corridor. This can be provided by 5 anchor nodes or 12 distributed-network (‘corridor-wide’) sites. A high-level description of each hub and network site is shown in Table 6. The preceding sub-sections describe candidate locations for siting anchor nodes and distributed network sites. Subsequently, unit, site, and overall corridor specification and cost estimates are discussed in the system-wide economic and environmental performance comparison, in Annexure. The estimated cost per station under the anchor node strategy ranges between INR 53 million and INR~90 million (INR 5.3 crore and INR 9 crore), whereas the corresponding estimates in the network strategy for EVCI build-out were between INR 140 and INR 230.

Siting for 30 per cent ZET

Anchor node ‘hubs’ strategy or Nodal ‘distributed network’ strategy moderates EVCI gaps and requirements in northern and middle stretches up to Gadpuri, whilst targeting severe deficits in southern stretches from Karman onwards, and especially in the Mahuvan<>Agra stretch

Figure 16. Five potential locations for the anchor node ‘hub’ EVCI build-out strategy along the NH44 corridor

Anchor Node or ‘Hub’ — Multiple EV chargers currently exist and are operational along the Delhi–Gadpuri toll stretch, primarily located near the Okhla industrial area, a key freight hub in Delhi. This section of the corridor serves as a critical link that connects Faridabad with Delhi. This stretch is characterised by a high density of offices, industries, warehouses and logistics facilities. A substantial amount of capital is required to set up a new charging station hub across this stretch due to the high cost of land. Based on the site suitability scores on the Gadpuri to Agra stretch, the proposed top five candidate anchor nodes or hubs are presented in Figure 16. 

These locations are strategically positioned at the city entry and exit points (e.g., Palwal, Hodal, Mathura, Vrindavan, Agra), ensuring high demand due to ongoing commercial activities. These five locations align spatially with the trip O-D, as illustrated in Figure 8. The identified anchor nodes, viewed together with the regional industrial and commercial activities, fulfil the quantum and gradient of EVCI requirements—moderate gaps in northern and middle stretches (sites A and B) whilst targeting severe deficits in southern stretches (sites C, D and E), especially Mahuvan<>Agra. Each of these sites is briefly described below.

• Site A (Palwal exit), situated almost midway between the Gadpuri and Karman toll plazas, is a prime terminal or transhipment node considering its proximity to ICD Palwal, a central logistics park and a host of auto and auto-allied activities, such as dealers, repair shops, and general auto parts and components manufacturers (Kongsberg, Agro). Furthermore, Site A is near a formally designated NHAI wayside-amenities (WSA) location.
• Site B (Hodal entrance towards Delhi) is tailored to serve bidirectional traffic by virtue of being in the vicinity of logistics, industrial (Tata and Eicher) and MSME clusters, such as Okhla and Faridabad, as well as truck stops and dhabas, increasing the likelihood of consistent traffic and utilisation.
• Site C near Mathura is located at a vital junction lined with metal- and plumbing-related industries and a major IOCL refinery. From here onwards, the highest truck volume enters the completely unserved Mahuvan<>Agra stretch.
• Site D is positioned to best cater to regional and inter-city freight from the heartland into Agra, feeding in and out of its periphery, which generates freight activity from transporting dairy products, small-scale industries, handicrafts, agri-markets or APMC mandis, and wholesale distribution networks.
• Site E near Agra is proximal to wholesale markets, rail-road transfer terminals, and warehousing clusters, making it well-suited for long-haul HDTs and deepcharging. This site is also very close to the upcoming integrated manufacturing cluster (IMC) along the inner ring road.

Siting for 30 per cent ZET

Distributed network or ‘corridor-wide’ strategy mitigates range and charging anxiety whilst improving redundancy, coverage and accessibility.

Figure 17. Potential sites for a distributed network EVCI build-out strategy along the NH44 corridor

Distributed network or ‘corridor-wide’— On the Delhi to Gadpuri stretch, the existing chargers are functional and well distributed. Rather than new installations, an upgrade is recommended. The Gadpuri to Agra stretch constitutes approximately two-thirds of the total corridor length, and most existing chargers are non-functional. New charging stations are required to ensure adequate coverage. Based on the infrastructure suitability analysis, locations with a suitability score above 0.45 have been selected. Four locations per stretch have been identified for new installations. Between the Gadpuri and Karman toll plazas, existing chargers are located near proposed sites 3 and 4. In such cases, the existing chargers can be retained instead of installing new ones. Notably, proposed location 3 is an NHAI wayside-amenities site, making it an ideal candidate for new EVCI installation. 

Only two chargers are operational between the Karman and Mahuvan toll plazas, which is also the longest stretch of the corridor. Given the spatial distribution, only the existing chargers between proposed sites 5 and 6 can be considered, as others are either not working or very close to existing ones. In total, five locations are recommended (four proposed + one existing).

Proposed charger locations have been strategically placed at entry and exit points of towns and cities, major intersections, industrial and commercial hubs, and designated rest stops, ensuring accessibility and alignment with traffic flow. Several warehouses, mandis, and factories are located near these sites. In the Mahuvan<>Agra stretch, as there are no existing chargers, proposed charging locations based on the suitability map are positioned towards and closer to Agra.

• Northern section (Locations 1-3: Delhi–Badarpur–Gadpuri region) — Higherfrequency, lower-volume MDT operations outbound from Delhi/NCR’s manufacturing zones. This segment experiences relatively balanced traffic flow that is lower compared to southern stretches, partly due to alternative routing options via the Eastern Peripheral Expressway diverting a portion of HDTs away from NH44.
• Central section (Locations 4-7: Gadpuri—Karman—northern Mahuvan) — This region includes Palwal and Hodal, which serve as critical logistics and freight consolidation zones. Hodal features several ancillary automotive suppliers, as well as dairy processing and Subzi mandi. These facilities generate significant refrigerated truck traffic transporting temperature-sensitive dairy products and time-bound perishables in both directions (Delhi and Agra).
• Southern section (Locations 8-12: southern Mahuvan–Agra) — This stretch experiences substantial tanker traffic due to the Mathura Refinery, locus for regional agricultural freight, particularly dairy and processed food products requiring refrigerated transport (e.g. Nandvan mega food park). A higher proportion of HDTs carry petroleum products alongside a mixed profile of MDTs/HDTs moving agricultural and consumption goods, as well as manufacturing inputs.

System-wide performance indicators

Hub-based strategy offers a greater, it requires an additional CAPEX between INR 530 lakhs– INR ~900 lakhs, whereas the corresponding estimates in the network strategy for EVCI build-out were contingent on parallel grid decarbonisation.

Table 7. Anchor node vs distributed network for EVCI rollout: A cost comparison (in INR lakh)

Source: (BEE 2024; CEA 2023, n.d.; MHI 2024; Tata n.d.; TSSPDCL 2023; UHBVNL 2023; UPERC 2022; ISGF 2024)
Notes: Detailed cost benchmarks and compiled data in Annexure.

Unit costs, site-level specification and the overall CAPEX to meet the charging demand are presented in Table 7. The CAPEX for five hubs across NH44 costs between ~INR 720 million (INR 72 crore) and INR 780 million (INR 78 crore), depending on whether 240-kW or 350-kW chargers are chosen. In the network strategy EVCI roll-out, the CAPEX varies between INR 560 million (INR 56 crore) and INR 660 million (INR 66 crore) for 350-kW and 500-kW chargers, respectively. The economics of network strategy versus hubs can be seen in Table 7, particularly with higher-capacity chargers, leading to 5 per cent to 25 per cent overall CAPEX reductions.

It should be noted that the cost figures presented are approximations compiled from extant studies and industry benchmarks rather than detailed site-specific quotations, such as the bill of quantities/materials (BoQ/M). Land acquisition costs have been deliberately excluded due to varying rates across the Delhi–Agra corridor, with reliable land parcel rate data unavailable. Nevertheless, the CAPEX estimated in Table 7 is based on the best available data and is plausible. Though the absolute unit, site or system-wide costs may vary during the realworld implementation phase, cost differentials provide directionally cogent inferences.

Potential fuel cost reductions and thereby GHG benefits of ZET transition have been studied from various angles — sectoral benefits of ~3–4 gigatonnes of cumulative CO2 reduction through 2050 (Raina 2023), 9–35 per cent GHG reduction depending on MHDT segment (Abhyankar 2022), 840 billion litres of diesel savings by 2050 (GoI 2023). and up to 80 per cent less GHG savings over vehicle’s lifetime (Yadav 2024), to name a few. An overarching theme governing such studies and their respective findings is the importance of decarbonising the electricity grid. Renewables account for 45 per cent of India's existing capacity of ~450 GW (GoI 2024a). However, 70 per cent of the generation is thermal — coal and lignite (GoI 2024a). The average electricity grid emissions factor is 0.75 kg CO2 /kWh for the vintage year 2024 and varies between 0.55 and 0.96 kg CO2 per kWh (CEA 2024).

Time-of-day, charging location, charging duration, regional grid mix, season and meteorological influence, renewable energy integration and coordination with charging and subsequently effect the potential for reducing charging emissions. Nascency of ZETs, limited to the non-availability of highly resolved spatio-temporal charging patterns of representative ZETs or utilisation data from compatible public chargers (rated at least CCS2, 120 kW), further limits the ability to conduct spatially and temporally resolved granular analysis. In light of these challenges, a stylised reduced-form approach has been adopted to evaluate whether and to what extent the EVCI strategy chosen has any bearing on charging-related emissions. This approach is outlined below.

First, using ISRO's solar calculator, at each of the chosen hub and network sites, the PV potential was evaluated. Average PV output was estimated to be 7.65 kWh per m2 per day. Referring to Annexure Table A22, there was little to no variation in PV potential across all 17 locations (5 hubs and 12 network sites) in the entire corridor. Structured scenario analysis was performed to determine fuel consumption across all combinations of low/medium/ high (L/M/H) of select parameters — distance, energy consumption, fuel efficiency, and PV utilisation (see Annexure Table A24). A few other assumptions were made regarding the realistic share of area available for PV installation on-site (40 per cent), and the incremental area for PV in anchor nodes (25 per cent) over distributed network sites. The area required per charger is based on MoHUA guidelines for public EVCI (15 × 7 = 105 m2 per charger) (MoHUA 2015). Five-number summaries across all modelled scenarios (81 in all) are presented in Annexure Table A25. Annual energy demand (under the 30 per cent ZET or N = 1,000 ZETs) varied from 56,250 kWh to 2,18,750 kWh. Diesel consumption and emissions ranged from 18,750 to 62,500 litres, and 51,563 kg CO₂ to 1,71,875 kg CO₂, respectively. Charging-related emissions lie between 42 t and 164 t CO₂.

Based on the PV generated, PV capacity (in kW) and battery energy storage system (BESS) back-up (1-hour of the same capacity as the PV and 4-hour options) were estimated (See Annexure Table A26). Based on publicly available data and industry estimates, the additional CAPEX for PV, BESS and the integrated PV+BESS were then calculated (Refer Table A18 and Table A26). The anchor node strategy can generate 2–4 times more PV than the network strategy. Simultaneously, it requires an additional CAPEX between INR 53 million to INR 90 million, whereas the corresponding estimates in the network strategy for EVCI build-out are between INR 14 million to INR 23 million Furthermore, both the anchor node and distributed network strategy for EVCI roll-out with PV+BESS can fully offset the charging-related emissions. This stylised and reduced form scenario analysis demonstrates the trade-offs of integrating PV+BESS for emission-neutral ZET charging.

Augmentation is a resource-effective option to reduce CAPEX and set-up time. Typically, it involves increasing the cable amperage to handle higher loads, reconductoring, replacing ageing transformers with higher-capacity units, and installing intermediate poles or aerial bunched cables to minimise line losses. This is suited for retrofitting areas already served by 11 kV networks, where incremental enhancements could suffice, are technically feasible and economically prudent. Co-location is another use case for augmentation. Charging stations situated at existing fuel stations, logistics hubs, or otherwise, are already equipped with commercial and industrial-grade electrical connections that could be reinforced.

Figure 18. Barriers to ZET adoption

Source: Field survey conducted by CEEW, 2025
Note: Respondents were asked to select up to 2. Heatmap shows the combined response in matrix form. 

ZET adoption perceptions of LSPs almost three-fourths-fourths remain opposed or uncertain due to high upfront costs, charging infrastructure gaps and the misalignment between the current ZET product landscape and their preferences—longer-range (300-500 km), extended warranty (8+ years), battery swappable variants and asset-light ownership.

This add-on component of the overall data collection exercise, previously clarified in the ‘Data and Methods’ section, surveyed LSPs—a key stakeholder group in the trucking ecosystem. This survey of 100 LSPs revealed key insights into their fleet composition, operational patterns, and perceptions regarding ZET adoption expectations—policy, technology, and business models. The aim was to learn ground-truth logistics needs and fleet preferences with the status quo, and identify the policy levers needed to catalyse ZET penetration. Descriptive and distributional data summaries are provided in the Annexure.

The findings revealed that 46 per cent of LSPs are unwilling to electrify their fleets, 19 per cent favour EV adoption, and 35 per cent remain uncertain about the transition. The respondents highlighted several barriers, with the upfront cost and limited availability of EV freight vehicle models emerging as the primary concern. Other barriers to EV transition in freight include long charging durations, range anxiety, lack of adequate charging infrastructure, and insufficient government incentives. The heatmap in Figure 18 shows that the high upfront cost and limited model offerings are the dominant—and perhaps mutually reinforcing—barriers to ZET adoption. Technological reliability and usability challenges manifest as a pressing hurdle highlighted by the overlaps between range anxiety, charging concerns, and model offerings.

The survey captured key requirements of LSPs regarding fleet electrification—the candidate vehicle they are most likely to replace with an EV, and its desired specifications. This includes battery capacity, range and warranty, axle configuration, ownership model, maintenance expectations, and charging preferences. To better understand the actual needs, the study focuses on the responses of LSPs who showed their willingness for EV transition (Yes) and are uncertain about it (Maybe). Together, these segments account for 54 per cent of the sample. Regarding battery specifications, 40 per cent of the LSPs prefer a battery capacity of 250–500 kWh, with an optimal driving range of 300–500 km. Their battery warranty expectations almost equally favour 8 years/500,000 km, or 10 years/1 million km, ensuring long-term reliability, cost-effectiveness, and assuaging concerns about battery replacement.

Under vehicle specifications, 51 per cent of the LSPs expressed their willingness for HDT (above 12 tonnes GVW) in a 6 X 4 axle configuration, reflecting the industry’s demand for high-capacity transport solutions. In terms of body type, refrigerated trucks emerged as the most preferred option (64 out of 100 LSPs), followed by a similar share of preference towards flat beds and tankers. Regarding ownership models, leasing and rental models have gained traction among LSPs, as they eliminate the need for high upfront investments and mitigate the risks associated with direct ownership. Additionally, for vehicle maintenance, 54 per cent of the LSPs preferred third-party service providers, reducing their operational burden and ensuring efficient fleet management. Interestingly, a quarter of the respondents expressed interest in vehicle usage or a per-km-based annual maintenance contract (AMC).

Charging infrastructure inadequacy was one of the prominent concerns among the LSPs surveyed. While ultra-fast charging was preferred (150–350 kW), a notable demand for battery swapping was also observed, and approximately 40 per cent showed a willingness to explore this option. The LSPs showed consensus (72 per cent) on a 2–4 hour charging time span, striking a balance between current operational efficiency and downtime. In terms of access, logistics hub-based charging was the most preferred approach, followed by highway public charging, ensuring convenience and minimal disruption to existing supply chain operations.

Zero emission truck transition requires system wide efforts and initiatives that involves multiple stakeholders across the value chain. In the context of charging infrastructure and considering the nascency of ZET penetration below are the summary 10 key recommendations for accerlerating ZET transition (Table 8).

Table 8: Summary of key policy recommendations and interventions

Source: Authors’ analysis
Note: For each action item, the relevant stakeholder group(s) are indicated in blue.