Electric construction machines have matured quickly, but deployments on remote sites often fail for a simpler reason: power arrives late, intermittently, or at the wrong magnitude. Unlike depot-based fleets (buses, vans) with permanent grid-tied infrastructure, construction sites are transient, access-constrained, and subject to highly variable duty cycles. A site may operate for months and then relocate, which makes “build permanent infrastructure” an uneconomic default in many cases.
The result is a widening gap between (a) what the machine can do and (b) what the site can reliably supply. Fast charging compresses demand: one excavator may accept hundreds of kilowatts, so a lunch-window “all plug in” event can produce megawatt-class peaks. Sizing infrastructure against average energy consumption is therefore a common and costly error. This paper provides a practical framework for thinking about temporary power as a portable microgrid system that can be sized, deployed, and operated as a jobsite asset.
Contributions and practical outputs of this paper include: (i) a clear sizing logic for peak power versus daily energy, (ii) a modular “mobile microgrid” reference architecture for construction, and (iii) an operations-centric checklist for mitigating downtime and safety risk.
Remote construction sites are governed by logistics: fuel delivery, security, maintenance, and schedule risk. Diesel has historically won because it is easy to transport and refuel, even when efficiency is poor. However, diesel logistics impose non-obvious costs: dispatch coordination, theft exposure, and frequent generator maintenance—especially when generators are oversized and run at low load factors.
Cargo and fuel theft have become a material operational risk for remote sites and supply chains. For example, a Verisk CargoNet analysis reported record levels of cargo theft in 2024 and noted that the average value per theft increased to roughly $202k.6 While those data refer to cargo broadly (not only construction fuel), they illustrate why “delivered energy” can carry a meaningful security and handling premium on remote projects.
Electricity logistics impose a different set of risks: undersized charging infrastructure causes breakers to trip, cables to overheat, and machines to queue or sit idle. Therefore, the core problem is not whether the vehicle battery is large enough in theory, but whether the site can deliver the required energy throughput and peak charging power with high reliability under real-world constraints (weather, dust, vibration, and variable production tempo).
Power (kW) is the instantaneous rate at which energy is delivered or consumed; energy (kWh) is the accumulated quantity over time. On electric jobsites, energy determines whether the fleet can complete a shift, but power determines whether the fleet can be recharged during narrow charging windows. Because fast charging events are spiky, peak power frequently becomes the binding constraint—even when total daily energy is modest.
Illustration: the Volvo EC230 Electric has a 450 kWh battery capacity and can use a 250 kW DC fast charger. 1,2 If a site attempts to recharge multiple machines simultaneously during a one-hour break, the microgrid must supply the coincident peak immediately. This is why inverter sizing is typically driven by peak concurrency, not by the daily kWh total.
A jobsite “mobile microgrid” can be treated as a modular, self-contained power plant that is designed for rapid deployment and relocation. A practical architecture includes four blocks:
Hybrid operation (generator plus battery) is often a practical intermediate: the generator can run near an efficient operating point to recharge the BESS, while the BESS handles fast-changing and peak loads. Hybrid microgrid literature finds that combining generators with storage can reduce fuel use and life-cycle costs in many settings.15 Jobsite deployments have also reported substantial reductions in generator runtime when batteries buffer variable loads.13
Sizing requires two parallel calculations: (A) daily usable energy (kWh) and (B) coincident peak power (kW). A robust design also applies derating multipliers for temperature, aging, and operational uncertainty.
Table 1. Key sizing variables and derived metrics for mobile microgrid planning.
List each machine type, battery capacity, expected depth-of-discharge per shift, and intended charging windows. Use conservative assumptions if duty cycles are uncertain.
For each machine i, estimate E_i = BatteryCapacity_i × DepthOfDischarge_i. Sum across the fleet to obtain E_day.
Define the maximum number of machines that may charge simultaneously. Compute P_peak = Σ ChargerPower_j over concurrently charging machines.
Apply multipliers for cold-weather capacity loss, battery aging, and contingency reserve. A common planning approach is to size usable energy 20–30% above E_day.
Select inverter rating P_inv ≥ P_peak with headroom for startup transients and future fleet growth. Verify continuous and short-duration overload ratings.
Size the average recharge power as P_avg ≈ E_day / T_recharge, where T_recharge is the available recharge time (e.g., overnight). Ensure fuel logistics and redundancy match schedule risk. Validate layout, cable routing, exclusion zones, grounding, and emergency shutdown; confirm siting requirements and test-report availability (e.g., UL 9540A) for the selected BESS.
Consider a high-intensity grading and loading shift with the following equipment mix:
Energy demand (kWh/day). Assume each machine uses 80% of its battery capacity during the shift. Then estimated daily energy throughput is:
E_day ≈ (2 × 450 × 0.8) + (423 × 0.8) + (72.6 × 0.8) ≈ 1,115 kWh (~1.1 MWh).
Coincident peak power (kW). Assume a lunch-break event where the fleet attempts to fast charge concurrently. Using 250 kW per excavator 1,2, ~300 kW for the loader (fast-charge class) 3, and ~60 kW for the compact loader yields:
P_peak ≈ (2 × 250) + 300 + 60 ≈ 860 kW.
System sizing implication. Even though the daily energy is ~1.1 MWh, the site must support nearly 1 MW of instantaneous power to avoid charging queues. A practical design may target an inverter rating on the order of 1 MW, with usable BESS energy sized above the daily requirement after derating (for example, 1.4–1.8 MWh usable, depending on recharge-source limits and environmental conditions).
Cold weather and aging. Lithium-ion capacity and charging power can be materially affected by temperature and aging; planning reserves (e.g., +25%) help protect schedule reliability and reduce the probability of mid-shift energy shortfalls.
Hardware alone does not guarantee uptime. Operating strategy determines whether the site experiences smooth production or repeated charging bottlenecks.
Megawatt-class lithium-ion systems introduce hazards that must be managed explicitly: electric shock/arc-flash, thermal runaway, fire/explosion risk, and heavy-transport/rigging hazards. Jurisdictional requirements vary, but consensus standards are increasingly used by Authorities Having Jurisdiction (AHJs), insurers, and project owners.
NFPA 855 (Standard for the Installation of Stationary Energy Storage Systems) is a central reference for siting, separation, emergency planning, and documentation.7 Commentary on the 2026 edition indicates that Hazard Mitigation Analysis (HMA) becomes the default expectation for many ESS types, rather than an exception triggered only by stored-energy thresholds.8
UL 9540A test reports (thermal runaway and propagation performance) are frequently used to inform spacing, suppression, and performance-based design decisions.9 For mobile deployments, contractors should require test documentation from suppliers, define exclusion zones, and coordinate emergency procedures with the AHJ and responding fire department.
Operationally, roles should be separated: jobsite operators should perform daily visual inspections and basic functional checks; high-voltage service and enclosure access should be restricted to qualified personnel under an electrical safety program.
The economic question is not “Are batteries cheaper than diesel?” but “What is the delivered, reliable cost of energy and uptime?” Diesel generators convert fuel to electricity efficiently only under the right load factor, yet many field generators are oversized and operate at low load, which increases maintenance burden and can reduce fuel efficiency. 5
Battery costs continue to decline at the utility scale, and credible public sources provide cost and performance projections. 14 Mobile and ruggedized systems, however, typically carry premia for packaging, transportability, power electronics, and duty-cycle intensity. Therefore, the most favorable projects are those where (i) equipment utilization is high, (ii) charging windows can be orchestrated, (iii) generator runtime can be reduced materially through buffering (hybrid operation), and (iv) downtime penalties are high enough that reliability is valued explicitly. 13
A practical evaluation method is to compute a project-specific levelized cost of delivered energy (LCDE) for each option (diesel-only, generator-plus-battery hybrid, battery-only with weak grid, etc.) and then convert it into cost per unit of work using site productivity (e.g., cost per hour or cost per cubic yard moved). Sensitivity analysis on diesel price, site duration, and utilization is essential.
Mobile power units can serve more than EV charging. They can power site offices, lighting, tools, and temporary facilities, displacing small inefficient generators. Because they can be redeployed, contractors can treat mobile microgrids as fleet assets (similar to cranes): utilization across multiple projects is a key driver of economics. Research and industry demonstrations of mobile batteries emphasize the ability to shift charging to off-peak periods and provide flexible power where it is needed. 11,12
This paper is a planning framework. Actual designs must incorporate detailed load profiles, charger characteristics, interconnection requirements (if any), site geometry, ground conditions, and local code amendments. The worked example uses simplified duty-cycle assumptions and should be treated as a sizing illustration, not a stamped design.
Future work should incorporate empirical datasets from electrified jobsites to refine duty-cycle models, quantify charging concurrency under real workflows, and validate productivity outcomes. A second research direction is control optimization: integrating on-site renewables, predicting load spikes, and scheduling charging to minimize cost and downtime.
Remote construction electrification is primarily an energy-logistics problem. By separating energy (kWh) from power (kW), planners can avoid the common trap of sizing for averages when peaks drive failures. A mobile microgrid—battery storage, power electronics, rugged distribution, and control logic—offers a deployable pathway to support electric earthmoving fleets without waiting for permanent utility infrastructure. When paired with disciplined operating strategy and safety governance, off-grid electrification can reduce diesel logistics burden, improve jobsite flexibility, and enable the power quality and data infrastructure that future autonomy will require.
