Tree protection bonds and guarantees on construction sites

A tree protection bond is a sum of money held by the local planning authority, or secured by a guarantee, which is forfeited if protected trees are damaged during construction. It cannot be imposed by planning condition, because a condition requiring the payment of money fails the tests in national guidance. It has to come through a planning obligation under section 106 of the Town and Country Planning Act 1990, and that obligation has to satisfy regulation 122 of the Community Infrastructure Levy Regulations 2010.

So the practical question for a developer is narrower than it first looks. Not "can they ask for a tree protection bond", but "does this particular bond meet the statutory tests, and is the sum defensible".

Why an authority reaches for a guarantee

Tree officers ask for bonds because the enforcement options after the event are poor. Once a mature oak has had its root plate compacted by a tracked excavator, prosecution under section 210 for contravening a Tree Preservation Order recovers nothing for the site, and replacement planting under section 206 delivers a whip where a canopy used to be. A financial guarantee changes the incentive during the eight months when the damage actually happens.

Two situations attract a request almost automatically. One is a constrained site where the construction envelope sits hard against a root protection area and there is no realistic alternative haul route. The other is a scheme where the retained trees are the reason permission was granted at all, typically a garden site in a conservation area where the canopy carries the character of the street.

There is a third, less comfortable reason. Some authorities have been burned by a developer who treated the loss of a protected tree as a cost of doing business, and the bond is a response to that history rather than to your scheme. Worth knowing, because it changes how you negotiate.

How the sum gets calculated

There is no statutory formula. In practice authorities use one of three approaches, and the difference between them is large.

The first is an amenity valuation of the trees themselves, usually through the Capital Asset Value for Amenity Trees method, which produces a figure from stem diameter, condition, life expectancy, and an adjustment for how many people benefit from the tree. A single large street tree can value into five figures on that basis. The Helliwell system is the other established approach, producing a points-based amenity value.

The second is a replacement cost approach: what it would cost to plant and establish a semi-mature specimen of comparable effect, including several years of aftercare. This produces smaller numbers and is easier to argue about, because nursery stock and planting costs are quotable.

The third is a round sum per retained tree, which some authorities apply as a standard. It is the least defensible of the three and the easiest to challenge, because it is unrelated to the value of the particular trees.

Ask which method has been used, and ask for the working. An amenity valuation that has not been checked against the actual condition of the tree recorded in the Arboricultural Impact Assessment is not evidence, it is an assertion. If the survey grades a tree as category C with a remaining contribution of under ten years, a bond priced as if it were a category A veteran is out of scale.

Getting it into an agreement

Because a condition cannot require money, the bond sits in a section 106 agreement or unilateral undertaking. Regulation 122 requires the obligation to be necessary to make the development acceptable in planning terms, directly related to the development, and fairly and reasonably related in scale and kind. Each of those is a place to negotiate.

The drafting points that matter:

• the trigger for payment, which should be the start of works rather than the grant of permission, so money is not tied up during a funding period
• the form of the security, because a bank guarantee or a parent company guarantee is usually acceptable and costs far less in working capital than cash on deposit
• the definition of damage, which should refer to the tree protection plan and the arboricultural method statement rather than to a general standard of care
• the mechanism for determining whether damage has occurred, ideally an independent arboriculturist rather than the authority alone
• the release trigger and the long stop date

Get the release wrong and the money sits with the council for years. A sensible release is staged: part on completion of the substructure works, which is when most root damage occurs, and the balance 12 or 24 months after practical completion once the trees have gone through a full growing season and shown no dieback.

The alternatives worth proposing first

Most tree officers would rather have certainty on site than money in an account, and there is a package that often replaces a bond entirely.

Offer a condition requiring the protective fencing to be erected and inspected before any equipment or materials come on site. Under section 100ZA of the 1990 Act, a condition of that kind, which has to be discharged before development starts, cannot be imposed without the applicant's written agreement. Volunteering one is a real concession and reads as such.

Add an arboricultural method statement with a named supervising arboriculturist, a schedule of site visits tied to construction stages, and a requirement to submit signed monitoring reports to the authority within a set period of each visit. Where ground protection or a no-dig surface is proposed within a root protection area, the specification should be in the method statement and on the drawing, not described in a covering letter. The detail of that is set out in our note on retaining a tree through a construction programme.

Where the scheme includes new planting, a Landscape Strategy with a five year establishment and replacement clause gives the authority the outcome it wants without a cash mechanism.

That package answers the officer's real concern, which is that nobody will be watching the site in week six when the scaffolders need somewhere to stack.

When to push back

A bond is unreasonable where the trees are not protected, are of low quality, or sit far enough from the works that no credible damage pathway exists. It is also unreasonable where the sum bears no relationship to the trees, where it duplicates a remedy already secured by condition, or where it is set at a level that would fund the authority's wider tree programme rather than remedy harm on this site.

Test any request against three questions. Would the development be unacceptable in planning terms without it? Is it directly related to this development? Is the amount fairly and reasonably related in scale and kind to the harm it guards against? If the answer to any of them is no, say so in writing and ask for the obligation to be removed or the sum revised. Officers do move on this, particularly where you offer the supervision package instead.

Illustrative example: a 14 unit scheme retains three category B limes along the frontage, with a shared surface drive crossing the root protection area of one of them. The authority seeks a bond based on amenity valuation of all three trees. The developer offers a no-dig construction detail signed off by the project arboriculturist, monthly supervision reports, and a bond limited to the single tree actually at risk, released 18 months after completion. The officer accepts. Figures and outcome are illustrative only.

One honest uncertainty: practice varies widely between authorities, and there is no published national guidance specifically on tree bonds. What is settled is the law on conditions and on planning obligations, which is why the argument is nearly always won or lost on regulation 122 rather than on arboriculture.

This is general guidance rather than legal advice. Check the position with your local planning authority or a qualified professional before relying on it.

PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval. Browse the 66 report types to see what a tree-constrained site usually needs.