Solar Park Vegetation Management: Mowing, Grazing and Design

Grass and weeds in a utility-scale solar park come back every season for the 25 to 40 years the land is occupied ^[1], and every season they threaten yield, fire safety, access and equipment. Solar park vegetation management is therefore a performance and compliance function, not groundskeeping. My position is that vegetation cutting deserves the same planning discipline as module cleaning or inverter maintenance, and that the cheapest controls are fixed at the design stage: ground clearance, row spacing, fences, roads and cable routing.

Why solar park vegetation management matters

Shading and energy loss on the lower module edge

SolarPower Europe notes that vegetation management can be a significant share of a PV plant’s operating costs ^[1]. The first reason to pay it is yield. Cells are wired in groups protected by bypass diodes, so grass over the bottom cells can disable far more of a module than the shaded area suggests. SolarPower Europe calls the loss disproportionate to the shading and links persistent shading to hot spots and long-term module damage ^[1].

Field data agree. The aerial inspection provider Raptor Maps describes vegetation shading as “one of the most consistent and prevalent anomaly types identified year after year” ^[2]. In a field test, shading a few modules in one string also pushed a neighboring, unshaded string off its optimal operating point, and prolonged shading can cause burned cells, bypass diode failures and, in severe cases, fire ^[2].

Bifacial modules add a second exposure, because the rear side depends on ground-reflected light. Westerhof and co-authors found that grass partly retroreflects light back toward the sun. Treating grass as a diffuse reflector, as yield models commonly do, overestimated the yield of their modeled bifacial module by up to about 10%, with the largest error for long grass. At a 4.1 MWp vertical bifacial agri-PV plant in Germany, yield increased by up to 6.6% in the days after mowing, and the authors attribute only part of that rise to higher irradiance ^[3]. One vertical-layout site is not a design value, but it shows the direction of the effect.

I have not found a published curve linking grass height to annual energy loss for Central European plants. The working rule follows from the physics: nothing grows above the lowest module edge, and under bifacial rows the sward stays short all season.

Fire risk

Dry grass under modules is fuel next to energized equipment. The most specific benchmark I know comes from Victoria, where bushfire risk shapes solar farm design. Energy Safe Victoria expects grass near electrical equipment to be no taller than 100 mm during the declared fire danger period, minimal combustible material under modules, and non-combustible fire breaks, such as crushed rock, at least 10 m wide around substations, power conversion units, battery systems and the site perimeter ^[4].

Insurers see the same risk. Gallagher warns that grazing sheep can leave patches of dry grass and that clippings left on gravel roads can ignite, and recommends mechanically clearing dry grass from under the modules where sheep graze ^[11]. Central Europe does not face Australian bushfire conditions, but uncut grass that dries out under the modules in late summer is the same kind of fuel, so I use these figures as a reference for equipment zones.

Access, structures and cables

Vegetation also decides whether O&M crews can reach equipment, and SolarPower Europe’s maintenance schedule includes vegetation clearing on site and along paths. Woody growth into modules and structures causes direct mechanical damage ^[1]. Cutting carries its own risk: mowers can throw stones into modules or strike above-ground equipment ^[5], and an operator in a German survey reported module damage from stones thrown by a bar mower with an unguarded cutter ^[12].

In the same survey, sheep chewed cables under string inverters mounted too low beneath the modules, causing faults until the cables were sleeved in conduit and fenced off with mesh ^[12]. A Lancaster University review notes that cattle and horses can dislodge mounting systems and that pigs and goats can damage cables, which is why sheep are the usual choice ^[13]. Tall vegetation also hides hazards for staff: SolarPower Europe lists rodents, snakes, ticks and stinging insects among site risks ^[1].

Mowing, grazing, herbicides and seed mixes compared

Mechanical mowing

Tractor-mounted mowers or mulchers between rows, with trimmers under the modules, remain the default. They cope with any vegetation, but quality depends on the contractor and module damage is a real risk. Autonomous mowers struggle with ponding water, obstacles and woody plants ^[1]. The hidden cost is under the modules. In the Cornell comparison, conventional maintenance of an 8.9 ha site needed an estimated 344 labor hours per year: 36 hours of mowing and 308 hours of string trimming under the panels ^[6]. That ratio is the strongest argument I know for designing clearance and row geometry around the machine that will do the work.

Sheep grazing

In the same trial, 56 sheep rotationally grazed the site from May to November, vegetation never shaded the panels, and grazing needed 139 labor hours including travel, about 2.5 times fewer than mowing and trimming ^[6]. The limits are well documented: sheep leave thistles and woody plants, quality depends on the farming partner, and costs rise where graziers are scarce ^[1]. Sheep also graze very low, so the Lancaster review recommends late-season grazing, or low densities if summer grazing is needed ^[13]. In NREL’s cost dataset, some grazed sites still needed mowing ^[5].

Grazing needs infrastructure and a welfare plan. SolarPower Europe and BirdLife call for stocking in proportion to the grazing area and an assessment of shelter and watering needs, with local experts involved ^[14]. The Cornell authors recommend that developers provide wells, power outlets and predator-proof fencing, and offer multi-year contracts so farmers can invest ^[6].

Herbicides and their regulatory limits

Chemical control is quick, but SolarPower Europe lists erosion, runoff, permitting and community opposition among its drawbacks, and notes it is not legal everywhere ^[1]. Directive 2009/128/EC on the sustainable use of pesticides requires Member States to give priority to non-chemical methods wherever possible (Article 14), to reduce or eliminate applications on very permeable surfaces and infrastructure close to water, with buffer zones (Article 11), and to minimize or prohibit use in specific areas, including protected areas under the Birds and Habitats Directives (Article 12). National action plans set out how this applies ^[7], so check site rules locally.

My recommendation is to exclude broadcast herbicide from the O&M contract and allow only targeted treatment of invasive or legally controlled species, by a qualified operator, where mechanical control has failed. The Lancaster review agrees that spot spraying or weed wiping beats blanket spraying, because even selective products remove non-target plants ^[13].

Low-growing seed mixes

The design-led option is to seed a low-growing mix after construction so that less biomass has to be removed. SolarPower Europe groups this with high ground clearance as passive vegetation control: lower operating costs for higher capital and design costs ^[1]. NREL found per-activity costs at native or pollinator-friendly sites lower than at turfgrass sites, but total costs slightly higher, which the authors attribute to extra work in the first three to five years of establishment ^[5]. Budget for that period.

Cost and planning

Published cost data are thin, and very little is European. NREL’s analysis of 54 US sites found median vegetation O&M costs of about USD 455 per hectare per year (USD 0.94 per kWdc) for turfgrass and about USD 695 per hectare (USD 1.28 to 1.66 per kWdc) for native vegetation and sheep grazing, with wide ranges and a sample too small for statistically significant differences ^[5]. In the 2013 German survey, operators reported EUR 180 to 200 per hectare per mulching pass, while a 108 ha park maintained daily in summer by an employee cost about EUR 1,390 per hectare per year ^[12]. The spread is the lesson: frequency, under-module work and site geometry drive cost more than the headline method.

The NREL authors also describe how design shortcuts become O&M cost: low-capex features such as low array heights pass unnoticed through design and construction, are expensive to correct once built, and consume O&M money never included in the LCOE calculation ^[5]. Draft the vegetation management plan before the layout is frozen and hand it over with the as-built documentation. In the O&M contract, perimeter vegetation and environmental compensation measures must be specified explicitly ^[1].

How biodiversity requirements are changing the practice

The old default was a short, uniform sward. Biodiversity guidance moves the target toward less frequent, later and more varied cutting.

SolarPower Europe and BirdLife recommend avoiding herbicides where practicable, grazing management, native wildflower meadows from locally sourced seed, field margins 7 to 10 m wide, appropriate space between module rows and preventing the spread of invasive alien species ^[14]. The German bne review of monitoring reports attributes species-rich solar parks mainly to permanently extensive grassland management between rows, and found that wider sunny strips between rows increase the density of insects, reptiles and breeding birds ^[15]. In the UK, the Lancaster review of 185 articles recommends low-intensity, late-season grazing or cutting, varied vegetation structure and minimal agrochemicals, while accepting that strips under 1 m wide in front of the south-facing panel edge are cut as needed to prevent shading ^[13].

Germany has turned part of this into tender rules. For first-segment EEG solar tenders with bid dates from August 2024, bidders must commit to at least three of five criteria: modules covering at most 60% of the project area; mowing no more than twice a year with cuttings removed, or grazing at a site-appropriate stocking rate; wildlife corridors on sites with a side longer than 500 m, plus small-animal passage; habitat elements on at least 10% of the area; and no plant protection products or fertilizer ^[8].

These rules do not remove the need to cut; they change where and when. The approach I recommend is zoning: keep the strip under and in front of the modules, equipment zones and fire breaks short, and manage inter-row areas and margins for the permit’s biodiversity objectives, cutting in phases so that some areas are always left untouched ^[1].

Hungarian specifics: common ragweed

Hungarian sites carry an extra legal duty. Under Act XLVI of 2008 on the food chain and its official supervision, the land user must prevent common ragweed (Ambrosia artemisiifolia) from forming flower buds and keep the land in that state until the end of the growing season. If the duty is not met, the authority orders public-interest control and must impose a plant protection fine, and inspections can take place without prior notice ^[9].

Solar parks are exposed for a simple reason. A field experiment in central Hungary found that soil disturbance plus ragweed seed produced ragweed in every habitat tested, with high biomass in all except closed forest, and that disturbance alone was enough where recent farming had left seed in the soil seed bank ^[10]. A solar park on former arable land, with piling, trenching and vehicle movements across the site, fits that description. In the Hungarian EPC projects I manage, I treat ragweed as a construction-phase risk: re-establish a closed sward quickly, manage ragweed patches as a separate zone cut or treated before buds form, whatever the late-cutting regime elsewhere, and log every intervention as evidence of compliance.

Designing the site so it can be maintained

From a civil and site-works perspective, most vegetation decisions are made before the O&M contractor arrives.

• Ground clearance. Set the lowest module edge for the intended method. A German survey of 14 operators and 13 shepherds found a lower edge of at least 80 cm favorable for sheep, although 12 of the 22 grazed parks operated at 0.8 m or less ^[12]. SolarPower Europe lists high clearance as a passive control that lowers operating costs ^[1], and under-module trimming is where the labor goes ^[6]. NREL’s dataset, however, showed no large effect of panel height on total vegetation cost ^[5], so clearance matters most where the chosen method depends on it. Raising arrays adds cost and reduces energy density, so the decision belongs in the LCOE model, not a late value-engineering round ^[5]. For trackers, check clearance at maximum rotation.
• Grading and pile reveal. In one surveyed park, a design lower edge of 0.8 m fell to 0.3 m in places because of uneven ground ^[12]. Check finished levels and pile reveal tolerances against the minimum clearance, not the nominal one.
• Row spacing and row ends. Rows wide enough for the mower and headlands at row ends avoid the five-point turns the Cornell report describes ^[6], and wider sunny strips support biodiversity ^[15].
• Fences, gates and water. For grazing, plan internal paddocks, gates wide enough for livestock trailers and mowers, water points and shelter ^{[6, 14]}. Keep a mowable strip along the perimeter fence and reconcile small-animal passages with livestock containment ^[8].
• Roads and equipment pads. Gravel roads and crushed-rock pads around stations double as fire breaks if kept free of vegetation and clippings ^{[4, 11]}.
• Cable protection. Use the mounting structure as a cable route where possible, which also limits buried runs ^[14]. Leave no loose loops below the module edge, mount string inverters high enough, and sleeve exposed cables in conduit where livestock graze ^[12].

Practical tool: vegetation management plan template

This is how I structure a vegetation management plan. Zones, targets and timing must be adapted to the permit, the grazier and the site.

Survey before layout freeze: former land use and ragweed records; protected species and permit conditions; erosion-prone areas; local graziers and contractors.

KPIs: zero exceedances above the lowest module edge at inspections; shading anomalies in annual thermography; fire break compliance before the dry season; no flowering ragweed or authority notices; zero damage caused by maintenance; cost per hectare and per kWp; permit biodiversity indicators.

Contractor scope: method statement per zone with exclusion zones around cables and equipment; stone guards; response time for exceedances; herbicide only as approved spot treatment; grazing agreement covering stocking, welfare, water, shelter, liability and insurance; perimeter and compensation areas named explicitly ^[1]; dated photo logs.

Conclusion

Vegetation cutting in an EU solar park is not cosmetic. It protects yield, controls fuel next to live equipment, keeps the site accessible and, in Hungary, discharges a legal duty on ragweed. Biodiversity rules shift the effort toward zoning and timing rather than removing it. The cheapest plants to maintain are the ones designed for it, with clearance, rows, fences, roads and cables laid out for the machines and animals that will do the work for three decades.