Belgian Fleet Electrification Business Case

Belgian calibration — what this model assumes

Every default below is set for a transport operator based in Belgium. If the operator is based elsewhere, the diesel refund, home toll rate and subsidy ceilings change and the case must be recalibrated.

• Diesel professional refund €0.1913/L (2026) — Belgium refunds part of the excise on diesel to professional hauliers, so the model always uses the net pump price. The law set it slightly lower each year to 2026 (€0.1935, €0.1924, €0.1913) and sets no further step, so €0.1913 stays until the law changes. Cutting or scrapping it is one of the options for the 2027 federal budget (Federal Planning Bureau, June 2026), not a decision — either would strengthen the electric case.
• No professional diesel refund. The Netherlands does not refund diesel excise to hauliers, unlike Belgium (€0.1913/L in 2026), France, Spain, Italy and Hungary. A Dutch operator pays the full pump price, so the model uses the pump price as the net price. This is the single biggest structural difference from the Belgian case, and it works in favour of electric.
• This note depends on the country chosen in the top bar.
• Belgian diesel toll €0.285/km in Flanders from 1 July 2026 (Viapass / kilometerheffing, incl. the new CO₂ component, up from €0.204) vs €0.038/km for a zero-emission truck at the 2026 80%-discount rate — the largest single driver of the business case. That 1 July 2026 change raised the Flemish rate for this truck by about 40%, and trucks first registered before 1 July 2019 are automatically in CO₂ class 1, the most expensive (93% of vehicles are in it) — so an older diesel fleet is hit hardest. For a zero-emission truck over 4,25 t the external-cost component (air, noise and CO₂) is set to zero and only the infrastructure component is charged, with a confirmed declining reduction: 80% in 2026, 60% in 2027, 40% in 2028, 20% in 2029. On 1 January 2027 the diesel rate also rises with inflation, by about 4% (to roughly €0.296), while the electric rate roughly doubles to about €0.078 as its discount steps down from 80% to 60%; neither is published yet. The ZEV rate then follows the post-2029 scenario below.
• Dutch diesel toll €0.206/km against €0.040/km for a zero-emission truck, at the announced 2027 vrachtwagenheffing tariffs — the largest single driver of the business case. Since 1 July 2026 every truck over 3 500 kg pays per kilometre in the Netherlands; the distance charge replaced the time-based Eurovignet outright, so a Dutch fleet went from a vignette it barely noticed to a bill that scales with every kilometre driven. The zero-emission reduction is 75% by law, 81% in the tariff and, unlike the Belgian one, it is not on a declining schedule: tariffs are kept stable to 2029 and the discount is evaluated by 2030 (EU law caps it at 75% from mid-2031). A 22.3% rebate on all Dutch tariffs runs from 1/9 to 31/12/2026; it lowers diesel and electric equally and ends before the first modelled year, so it does not touch the P&L.
• This note depends on the country chosen in the top bar.
• The Dutch charge now matters too. Since 1 July 2026 every truck over 3 500 kg pays per kilometre in the Netherlands too, and at the announced 2027 rates that is €0.206/km diesel vs €0.040/km electric. With the default 18% Dutch share of tolled kilometres that saves about €2 700 per truck a year, the third-largest toll saving in the model after Flanders and Germany. A 22,3% rebate on all Dutch tariffs runs to 31/12/2026; it lowers diesel and electric equally, so it does not change the comparison, and it expires before the first modelled year.
• The Belgian kilometres are the volatile ones. In Flanders a diesel truck pays €0.285/km since 1 July 2026 against €0.038/km for zero-emission — an even wider gap than at home, and worth having on the roughly 15% of tolled kilometres the default mix puts in Belgium. But that Belgian discount declines on a fixed schedule (80% off the infrastructure charge in 2026, then 60%, 40%, 20%) and under current law it ends after 2029. Wallonia makes no distinction at all: there a zero-emission truck pays the diesel rate. So Belgian mileage helps today and helps a little less each year, which is the opposite of the German kilometres.
• This note depends on the country chosen in the top bar.
• Default profile — 100 000 km/yr, researched 6-zone mix. There is no official km-by-country split for Belgian trucks, so the default zone mix (VL 45 / WA 8 / BR 2 / NL 18 / FR 15 / DE 12) is a transparent, editable estimate for a Flemish cross-border operator on the main freight corridors, not the average Belgian truck (Eurostat: 70% of Belgian lorries' kilometres are domestic, and France is their biggest foreign market). A daily duty in this range is covered on a single overnight depot charge (Windrose R700 Gen-2 ~670 km loaded range, ~500–550 km realistic; the 2026 MAN eTGX shown at IAA quotes 660 km on the 4x2 and 720 km on the seven-pack 6x2 with 660 kWh net), so no en-route public charging is assumed and no public-charging premium is added. Cross-border work (Antwerp–Rotterdam, the Ruhr, Lille/Paris) is modelled with the editable zone split in the toll panel; Germany’s full ZEV exemption (€0/km through 30/6/2031) makes cross-border mileage toll-favourable for electric, not unfavourable.
• Default profile — 100 000 km/yr, an estimated 6-zone mix. There is no official kilometres-by-country split for Dutch trucks either, so the default (NL 55 / DE 20 / VL 12 / FR 10 / WA 2 / BR 1) is a transparent, editable estimate of a Dutch line-haul duty: most kilometres at home, then the Ruhr and the Antwerp–Rotterdam corridor. A daily duty in this range is covered on a single overnight depot charge (Windrose R700 ~670 km loaded, ~500–550 km realistic; the 2026 MAN eTGX quotes 660 km on the 4x2 and 720 km on the seven-pack 6x2), so no en-route public charging is assumed and no public-charging premium is added. Germany’s full zero-emission exemption through 30/6/2031 makes cross-border mileage toll-favourable for electric, not unfavourable.
• This note depends on the country chosen in the top bar.
• After 2029 the Belgian ZEV discount ends under current law. The amended Eurovignette Directive ((EU) 2025/2459, 26 November 2025) lets a member state exempt zero-emission trucks fully until 30/6/2031, but it is optional. Flanders chose not to: on 24 June 2026 its Parliament passed the CO₂ law without an amendment that would have kept electric trucks free until 30/6/2031, and Brussels has not moved either. So the discount still ends on 31/12/2029. From then on an electric truck pays the full infrastructure part plus a small external-cost charge of about €0.016/km in both regions, which the model leaves out. The default uses the conservative "cliff" (ZEV pays full infrastructure from 2030); the panel lets you switch to "free until 30/6/2031" or "permanent 25% of infra". This is the single largest toll uncertainty in the model.
• After 2029 the Belgian discount ends under current law — but it is not your biggest uncertainty. The amended Eurovignette Directive ((EU) 2025/2459, November 2025) lets a member state exempt zero-emission trucks fully until 30/6/2031, but it is optional and Flanders and Brussels have not used it. The default uses the conservative "cliff" (zero-emission pays full infrastructure from 2030); the panel lets you switch to "free until 30/6/2031" or "permanent 25% of infra". For a Dutch operator that switch only moves the Flemish and Brussels share of the mix. The uncertainty that would really move your case is a change to the Dutch reduction (75% in law, 81% in the tariff): tariffs are kept stable to 2029, the discount is evaluated by 2030, and EU law caps it at 75% from mid-2031.
• This note depends on the country chosen in the top bar.
• No subsidy is assumed. EU AFIF only funds public truck-charging hubs (at most 30% of works costs; call open to 6 October 2026), not private depots. The federal increased investment deduction takes 40% of the investment in a zero-emission truck or a truck charging point off taxable profit for purchases in 2026, for large and small companies alike (with a certificate; for trucks, applications since 1 April 2026) — roughly 10% of the price in cash. Since 1 January 2026 electric trucks are no longer on the Flemish Ecologiepremie+ list (VLAIO). Enter what you are awarded: under “Subsidy on trucks” it counts in year 1, under “Subsidy on infra” in years 2–4.
• No subsidy is assumed, and the Dutch stack is not the Belgian one. EU AFIF only funds public truck-charging hubs (at most 30% of works costs), not depots. What matters is Dutch: AanZET for the truck itself (round 2 from 29 September to 16 October 2026: 29% of the truck price for small firms, 21% for medium and 11.1% for large, at most €115,200, €83,200 and €43,900 per truck — pick it under “Subsidy on trucks” and the panel works it out for the chosen truck), SPRILA for logistics charging infrastructure (40% for SMEs, budget €122.5 million in 2026) and MIA as tax relief on the truck (since 2026 electric trucks no longer get Vamil, and heavy-vehicle charge points are off the list); budgets change from year to year, so nothing is assumed here. Enter what you are actually awarded: under “Subsidy on trucks” it counts in year 1, under “Subsidy on infra” in years 2–4. Check the current schemes before you count on them.
• This note depends on the country chosen in the top bar.
• Foreign mileage is charged at each zone's own year-indexed rate — set the per-zone split in the parameters, and see the toll table at the bottom for every one of the thirteen zones, its source, and the sensitivity notes.

Cost of capital / WACC — what it is and how to set it

The short version: your cost of capital is the yearly return you could earn on this money elsewhere at comparable risk — the hurdle the project has to clear. In finance it is also called the discount rate, because it is the rate at which future euros are "discounted" back to today: a euro you only receive in year 7 is worth less than a euro now, and this rate is exactly how much less. (Note: this has nothing to do with a price discount on buying a truck — it is purely about the time-value of money.) Future savings are divided by (1 + rate) for each year of waiting before being added into the NPV.

Why it matters here: it only affects the NPV — the one figure expressed in today's money. NPV discounts every future cashflow back to today at this rate, so a higher rate shrinks far-off savings more and NPV falls; a lower rate raises it. It also feeds the MIRR, which values the yearly savings at this same rate. It does not change the nominal "Total savings" (that is money-of-the-day, undiscounted). The test: if the MIRR exceeds your WACC, or equivalently NPV is positive, the project creates value above this hurdle.

How to set it for a Belgian transport operator:

• WACC (weighted average cost of capital) blends what you pay for debt and the return your equity expects. If most of the fleet is financed by a bank/lease, your WACC leans toward the loan rate.
• Typical 2026 range: 6-9% for a financed asset purchase. A truck lease at ~5-6% interest with modest equity lands near 7-8%. Use a higher figure (10-12%) if you want a stricter test or fund it mostly from equity.
• Rule of thumb: set it to your actual financing rate plus a small buffer. If you simply want to compare against a passive index, set it to that index's expected return (e.g. 7.5% for MSCI World) and a positive NPV then means the fleet beats the index in present-value terms.

Default here is 8.0% — a reasonable mid-point WACC for a financed Belgian heavy-fleet operator in 2026. Adjust the slider to match your own financing.

Charging infrastructure — how it is sized, costed & why these numbers

The key insight: a depot charges overnight, ~22:00–06:00 — an 8-hour window. A regional truck needs only ~420 kWh back, so it draws just ~52 kW averaged over the night; we size to ~65 kW/truck for headroom — a fraction of the truck's 400 kW maximum. Smart software staggers the trucks so the depot's simultaneous peak stays low. That is what makes it affordable: you size the grid and hardware to ~65 kW/truck, not to the nameplate.

One cabinet serves many trucks. The expensive part is the central power cabinet, shared across a row of trucks: a real unit such as the Huawei DS720 (720 kW) runs ~10 trucks at once, or more across the night since they don't all need full power at the same moment. Each truck has its own cheap dispenser, but the costly electronics are pooled — which is why the cost per truck falls steeply with fleet size.

Why this is a depot-only model. The trucks return to base and charge slowly overnight at the depot price (~€0.14–0.24/kWh, depending on the country). Public/fast charging is excluded — it's a separate, occasional, far pricier use case (≈€0.46/kWh, higher losses) that would distort a depot TCO. Megawatt charging (MCS) is also out of scope: it's for en-route top-ups on long-haul routes, with much heavier, costlier grid infrastructure and higher losses at megawatt power. Low-power overnight DC is both the cheapest and the lowest-loss path for return-to-base fleets.

Charging losses — why low power wins twice. Not every kWh reaches the battery: some is lost in AC/DC conversion, the transformer, cabling and standby. It depends on charging power — low-power overnight split DC ≈ 8% (≈5% conversion + 1% transformer + 1% cabling + 1% standby); high-power all-in-one ≈ 11–12% (see table below). This is a recurring cost. Example: 100 trucks, ~500 kWh/night, 300 nights, €0.16/kWh → 8% ≈ €209k/yr wasted vs ≈€327k/yr at 12% — about €120k/yr saved by staying low-power. Low-power overnight DC wins twice: lower CAPEX and lower loss.

What "turnkey" means. The single charging-infra figure in the TCO table is the all-in, ready-to-use price — everything needed to drive a charged truck out in the morning:

• Charging hardware — the power cabinets (the expensive electronics) and the dispensers/satellites at each parking bay.
• Civil works — concrete foundation pads, trenching, underground cabling, bollards and wheel-stops to protect the equipment.
• Grid connection & transformer — the medium-voltage (MV) connection and step-down transformer (see below).
• Engineering & commissioning — design, installation, testing, and the charging-management software set-up.

Because the fixed parts (grid, transformer, engineering) are shared across all trucks, the cost per truck falls as the fleet grows — from roughly €37k/truck at 20 trucks to ~€23k/truck at 200 trucks. That is why the figure recalculates whenever you change the fleet size.

How much hardware?

• ~1 power cabinet per 10 trucks and ~1 dispenser per truck — for pure overnight charging every truck dwells at once, so each needs its own plug. The dispensers are cheap; the power behind them is shared.
• Higher truck-to-charger ratios exist, but only with daytime turnover (NACFE recorded US Foods on 1 charger per 3 trucks). Overnight can't do that — all trucks park at once — so this model conservatively assumes 1 dispenser/truck. Trucks returning in waves need fewer, dropping the cost further.
• Grid: ≤5 trucks may fit the existing low-voltage connection; from ~20 trucks you need medium-voltage + transformer (Fluvius: >250 kVA is always MV).

The four charging systems — all DC. Split-architecture systems lose less energy and cost less; we prefer them and flag the higher-loss options:

"Auto · low-loss split DC" (the default) always picks a low-loss split-DC system — never a higher-loss all-in-one: Kempower for small/mid fleets (≤80), Huawei for large fleets. You can still override it to any vendor; the higher-loss options stay selectable for comparison, but the tool will always show you their extra loss cost. A cost band (optimistic / central / conservative) reflects how much real quotes vary.

Note on AC charging: this model is DC-only. Windrose trucks charge on DC (CCS / MCS), not AC — and DC is the right choice for a depot anyway: the AC→DC conversion happens once, in the shared cabinet, rather than being lost in every truck's on-board charger. AC depot charging is not considered.

Why we deliberately skip a battery buffer (BESS):

• A BESS (on-site battery) earns its keep by shaving short, sharp demand peaks — which happen with fast/public charging, where lots of energy is forced through a small connection in minutes.
• An overnight depot has the opposite profile: a long, flat, low-power load. With smart load-spreading there is almost no peak to shave, so a €0.5–2M battery would mostly sit idle. Independent modelling agrees: a 2025 Fraunhofer study found that for an overnight depot the investment in a battery storage system is not profitable regardless of capacity — its cost exceeds the savings — and NACFE/RMI warn that some fleets over-built storage they never used.
• So BESS is excluded from the base case. It only becomes worthwhile if (a) the grid connection is hard-capped below your managed peak, (b) you add daytime fast-charging, or (c) you co-locate large solar.

And the Belgian peak-demand charge (capaciteitstarief)? A Flemish business on medium voltage pays Fluvius partly for the power it books in advance (toegangsvermogen, €/kVA) and partly for its actual monthly peak (maandpiek, €/kW); going over the booked power costs 50% more. For naïve simultaneous charging this would be brutal — but because overnight software keeps the peak at ~65 kW/truck instead of 400 kW, both the booked power and the billable peak (and the charge) drop by roughly 4–6×. It becomes a normal, manageable operating cost rather than a deal-breaker — which is exactly why no battery buffer is needed to control it.

And the Dutch network tariff? A business connection here is billed on contracted transport capacity (gecontracteerd transportvermogen, €/kW), not on energy alone. Naive simultaneous charging would force that contract up sharply — but because overnight software keeps the peak at ~65 kW/truck instead of 400 kW, the capacity you have to contract and pay for drops by roughly 4–6×. In the Netherlands that counts twice: a smaller contract is not only cheaper, it is far more likely to be granted at all in a congested area. Which is exactly why no battery buffer is needed to control it.

This note depends on the country chosen in the top bar.

Turnkey ranges are planning-grade (±30%), triangulated from Alpitronic retail pricing (~€255/kW anchor), Smart Freight Centre, McKinsey, ICCT and 2025–2026 vendor data; high-power split systems (Huawei, Kempower, Siemens) are quote-only. These per-truck figures (≈€37k at 20 trucks falling to ~€23k at 200) sit in line with — and deliberately at the cautious end of — independent benchmarks: NACFE/RMI field data reports €14–19k per truck at scale, and the Fraunhofer/Transport & Environment depot-charging study (2025) puts infrastructure + grid connection at roughly €0.06/km for overnight (50 kW) charging. If anything this model is conservative, not optimistic. Get site-specific quotes before committing. Belgian grid context: Fluvius 2026 grid tariffs (Vlaamse Nutsregulator): booked power plus monthly peak, MV connection >250 kVA.

Grid capacity — the constraint that decides when you can start

Everything above prices the trucks and the chargers. This section is about the one thing money cannot buy quickly: a big enough electricity connection at the depot. The toll and fuel savings are real from the first kilometre, but you only collect them once the depot can actually charge. On congested parts of the network the connection, not the truck order, sets the start date — so it belongs in the business case as a schedule risk, not a line of cost.

How full is the network, in numbers

The Dutch figures are the ones that are published in detail, and they matter directly here: every default profile sends part of its kilometres to the Netherlands, and any operator with a Dutch depot or distribution centre is inside this problem.

What actually gets you around it

The good news is that the charging design in this model is already the low-congestion one. Overnight smart charging holds the depot at roughly 65 kW per truck instead of the truck's 400 kW maximum — a 6× smaller connection for the same energy — so the request you put in front of the grid operator is far more likely to fit inside existing capacity. Beyond that, the standard "behind-the-meter" toolkit applies:

• Smart charging and an energy management system — spread the load across the whole night window and cap the site peak. This is already assumed in the model and is the single cheapest measure.
• Rooftop solar — most depot roofs are large and empty. It reduces what you draw at the meter, though it produces little during the overnight window when trucks actually charge.
• Battery storage (BESS) — deliberately excluded from the base case because a flat overnight load does not need it. It becomes worthwhile precisely when the grid connection is hard-capped below your managed peak, which is what congestion does. If the grid operator offers less power than you need, price a battery back in.
• A flexible or time-bound connection contract — accepting curtailment at peak hours can move you up the queue considerably, and overnight charging is unusually well suited to it, because the hours you need are the hours the network is quietest.
• Phasing the fleet — size the first connection for the trucks you can actually charge, and treat later expansion as a separate application started early.

Belgium is not immune, but it is not the same picture. The detailed figures above are Dutch; Belgium publishes nothing directly comparable, and Fluvius itself says Flanders today faces expected rather than actual congestion. Where it expects the grid to fill up, Fluvius can give a new or heavier business connection as a temporary 'Fall-Back Flex' connection: you get the power you asked for, but Fluvius may ask you to limit it for a while, which overnight smart charging handles easily. The Belgian cost shows up mainly in the capacity tariff, which the overnight design already keeps low. Treat the Dutch numbers as the evidence for why to apply early and design for a modest connection, not as a forecast of your own Belgian lead time. Ask your grid operator for the capacity at your specific address before committing to a delivery date.

These figures are about your own grid. Everything above is Dutch data, and for a Dutch operator it is not a cautionary tale from next door — it is the connection you will be applying for. In August 2026, 45% of supply areas had a shortage with a waiting list, and a heavier logistics connection can take three to seven years. That is why the overnight design in this model matters more here than anywhere else: ~65 kW per truck is a connection you might actually be granted, where 400 kW per truck is not. Check the congestion map for your own postcode and talk to your network operator before you commit to a delivery date.

This note depends on the country chosen in the top bar.

Sources: national congestion statistics per August 2026 and the end-2025 application backlog; the 1/7/2026 rule extending waiting lists to small consumers; sector reporting on logistics connection lead times and behind-the-meter mitigation. Re-check before fixing any deployment date — these figures move every quarter.

Depreciation / residual value — how it is modelled

At the end of the contract the trucks still have resale value. The model captures the net residual benefit of the electric fleet versus the diesel fleet (electric resale minus the diesel resale you would also have had) as a one-off cashflow in the final year. It is period-specific — a 5-year-old truck is worth more than a 10-year-old one — and you choose one of three scenarios:

Why electric residuals are treated cautiously: used EVs have so far depreciated faster than comparable diesels — battery packs lose ~2-3% capacity per year, and each new truck generation brings better range and faster charging, making older models feel dated. Buyers also discount for perceived battery risk. The longer you hold (10 yr), the more battery degradation and obsolescence erode the residual, which is why the figures fall steeply with the period.

Battery replacement is not modelled separately; a major pack replacement within the term would materially change these figures. Sources: MDPI / Recharged EV-depreciation studies 2025-2026.