Sweden’s forestry industry moves enormous volumes of timber by road, and those hauls carry a heavy environmental and financial burden. Road transports of roundwood from forest landings to processing industries account for more than half of the CO2 emissions and roughly 40 percent of the operational costs of harvesting and transport. Because freshly harvested wood is dense and heavy, payload is usually limited by weight rather than volume, which makes gross vehicle weight a decisive lever for efficiency. A new simulation study from Värmland in Sweden now shows that simply buying bigger trucks is not enough: the value of larger and heavier vehicles depends critically on how well a haulier's fleet matches the road network it must operate on.

The study built a discrete event simulation model of roundwood transport in Värmland, where four wood products, pine sawlogs, spruce sawlogs, coniferous pulpwood and deciduous pulpwood, are delivered from 306 harvesting sites to two pulp mills and four sawmills. Each week about 5,500 cubic metres of wood, roughly 5,000 to 6,000 tonnes, is harvested from two to ten sites. The model, developed in ExtendSim software, assigns these transport tasks to fleets of vehicles with permitted gross vehicle weights between 64 and 90 tonnes over a simulated 44-week year, with each truck running 80 work hours per week in two eight-hour shifts, mirroring the annual working time of a Swedish log truck.

The study compared a reference fleet of standard 7-axle trucks against nine alternative fleet compositions, some mixing vehicle sizes and some consisting of a single type. These fleets were run on three versions of the road network: the current mix of roads allowing 64-tonne vehicles and roads allowing 74-tonne vehicles, a network in which a selection of bottleneck roads had been upgraded to the higher class, and a network where all public roads allowed 74 tonnes. Performance was measured with three indicators: distance driven per cubic metre of wood delivered, effective work time per cubic metre, and cost per cubic metre. Five replicate runs were made of every combination of fleet and network, with identical random number seeds within each replicate so that every scenario faced the same wood supply.

Swedish road regulations shape the vehicle options in a distinctive way. Roads are classified from BK1 to BK4 by bearing capacity, with BK1 permitting up to 64 tonnes and BK4 up to 74 tonnes depending on axle configuration. When heavier vehicles were permitted, axle loads were not increased, so reaching 74 tonnes requires extra axles: 8-axle combinations max out at 70 tonnes and 9-axle combinations at 74 tonnes. Those additional axles raise tare weight, purchase cost, rolling resistance and fuel burn per kilometre. On roads restricted to 64 tonnes, a 9-axle truck can actually carry less than a standard 7-axle truck, because its heavier unladen weight eats into the permitted load. The study also included a fictive road class, BK4+, assigned to BK4 roads at least 3.5 metres wide, on which an 11-axle, 90-tonne combination was assumed to be permitted with a special permit.

The results reveal a sobering picture for hauliers hoping that heavier trucks will automatically cut costs. On the current road network, no fleet composition performed significantly better than the 7-axle reference fleet in terms of cost per cubic metre. Only fleets containing at least one 11-axle 90-tonne vehicle significantly reduced distance driven per cubic metre, and for work time per cubic metre the improvements extended to fleets FC4 and FC5 as well. Uniform fleets of only 8-axle or only 9-axle trucks did not significantly improve any performance indicator on the current network. The reason lies in utilisation: in the simulations, more than half of the loads carried by 8-axle trucks were on roads that limited them to 64 tonnes, wasting their extra capacity, while 9-axle trucks, though better placed, still suffered from their higher unladen weight on restricted roads.

Removing a handful of identified bottleneck roads, reclassifying them from the lower to the higher bearing class, produced no significant improvement in any performance indicator across the ten scenarios. The authors attribute this to two factors: other parts of the network still forced transports onto 64-tonne routes, and acceptable 74-tonne detours were often already available that were not much longer. The picture changed dramatically when the entire road network was upgraded to allow 74 tonnes. Under that scenario, a uniform fleet of 9-axle trucks significantly improved all three performance indicators compared with the 7-axle reference on the current network, and the model estimated that overall transport costs would be reduced by about four percent and fuel consumption by roughly three percent. Even so, the authors note that this gain was only about two percent better than the most efficient mixed fleet operating on the current network, raising the question of whether the societal cost of upgrading every road would be justified.

Fuel consumption figures across the fleets told a similar story. On the current road network, calculated fuel use per cubic metre was approximately similar for all fleets except the all-8-axle and all-9-axle fleets, where low utilisation on restricted roads drove consumption up. When all roads were upgraded to the higher class, calculated fuel consumption was equal to or slightly lower than the reference in every scenario, pointing to real potential for savings from network upgrades. The authors caution that these fuel figures derive from the input data of the truck costing calculations and should be read as approximations rather than precise measurements.

Diesel was priced at 16.4 kronor per litre including AdBlue. For 11-axle vehicles, which are too long to load with their own cranes, a separate loading cost of 17 kronor per tonne was added to cover the extra crane vehicle. Even with these accounting assumptions, fleets containing 11-axle vehicles did not reduce transport costs relative to the reference fleet, although they did cut work time per cubic metre. The authors suggest these very large vehicles may be better deployed as shuttles between sawmills and pulp mills, where more efficient loading options exist than on small forest landings.

Several model design choices matter for interpreting the findings. Products were always delivered to the closest accepting industry, which minimises distance but does not reflect real-world supply chain optimisation, where contract volumes and the need for even delivery flows can send wood to more distant mills. The case area is well studied and representative of regional transport for a wood procurement company, but it is small: about 296,000 cubic metres across 306 landings, compared with the 72.4 million cubic metres of Swedish roundwood transported in 2022 from more than 280,000 landings to roughly 330 recipient industries. The reference scenario also differs from earlier studies by using the current road network rather than assuming all roads are restricted to the old 60-tonne limit, giving the 7-axle baseline a higher average payload and making significant differences harder to detect.

The study’s conclusions are pragmatic rather than revolutionary. Adapting fleet composition to the available road network is beneficial from work time, fuel consumption and cost perspectives, and it is important to balance the fleet so it matches the roads. Heavier vehicles are worth striving for, but the network restricts how many are useful, and using them on inadequate roads brings losses in investment cost, fuel consumption and payload. For the current road network in the case studied, a fleet of 7-axle and 9-axle vehicles is preferable if transports can be planned in advance, while 8-axle trucks remain versatile options for hauliers facing variable landing availability throughout the year. Upgrading roads from the 64-tonne to the 74-tonne class decreases work time, distance travelled and costs per cubic metre, but the benefit will probably be smaller for less used roads, and because much of the peripheral network consists of private forest roads, entire routes must be available for heavier trucks to pay off. The authors suggest that simulation tools of this kind could help hauliers find suitable fleet compositions for their own needs, and that further research is needed to determine where road upgrades are economically viable.

Subject of Research: Simulation of truck fleet composition and road bearing capacity in Swedish forestry transport

Article Title: Impact of fleet composition and road network on efficient use of larger and heavier trucks in forestry transport

Article References: Eliasson, L., Enström, J., Jönsson, P., Parklund, T., & Pernestål, A. (2026). Impact of fleet composition and road network on efficient use of larger and heavier trucks in forestry transport. European Transport Research Review, 18(1), Article 79. https://doi.org/10.1186/s12544-026-00840-2

Image Credits: AI Generated

DOI: 10.1186/s12544-026-00840-2

Keywords: forestry transport, roundwood logistics, gross vehicle weight, discrete event simulation, road bearing capacity, fleet composition, high capacity trucks, fuel consumption, transport costs, Sweden, Värmland, road upgrades