Building on Difficult Sites in Cumbria

The cheapest point at which to solve a difficult site's problems is before you own it. A plot that looks like a bargain because of its slope, its access or its isolation is usually priced that way because the market has already done the sums on what it will cost to build there. Working through those sums before exchanging contracts, rather than after, is the single most valuable thing a buyer can do.

Start with access. Walk or drive the actual route a concrete wagon, muck-away lorry and delivery vehicle would need to take, not the route shown on a site plan. Note width pinch points, weight-restricted bridges, passing places, turning space and any section that a heavy vehicle simply could not use. A plot with no vehicular access for standard construction traffic is not automatically undevelopable, but it does mean smaller loads, more trips, tracked plant and self-delivery — all of which cost time and money that a straightforward site would not.

Check the topography properly, ideally with a level survey rather than a guess from standing on the ground. A slope that looks gentle from the road can conceal a much steeper drop across the building footprint, and the difference between a modest cut-and-fill and a site needing significant retaining structures is exactly the kind of thing that is invisible until it is measured.

Establish the drainage position before anything else is committed. Where does foul water go? Is there a mains sewer nearby, or will a treatment plant and drainage field be needed? Is there a watercourse available for surface water discharge, or will the ground need to absorb it via a soakaway? On low-permeability or high-rainfall ground, percolation testing may show that a soakaway will not work at all, which changes both the drainage design and the space it needs on the plot.

Ask about services. Water, power and telecoms connections to a remote or elevated plot can involve long private runs, wayleaves across other people's land, and lead times measured in months rather than weeks. Get an actual quotation from the relevant utility or a competent installer before assuming a connection is straightforward.

Finally, check the planning position specific to that plot rather than a general assumption about the area. Any outline or full permission attached to the land may carry conditions on access, drainage, materials, levels or ecology that were negotiated for a particular scheme and may not transfer cleanly to a different design. Confirm the current position with the local planning authority, and with the relevant National Park Authority where the plot falls within a designated area, before treating the plot as buildable on your terms.

Every decision that follows on a difficult site — cut-and-fill balance, retaining wall height, drainage falls, foundation depth — depends on knowing the existing levels accurately. Guessing from a site visit or working from a drawing based on an old Ordnance Survey contour map is not good enough on ground that is doing anything more than sitting flat.

A topographic survey records spot levels across the site on a grid, together with boundary positions, existing structures, trees, banks, watercourses and any visible services. On a sloping or irregular plot this is what allows a designer to place the building at the level that minimises earthworks, rather than at a level chosen for the view or the drawing board and corrected expensively on site.

Once a survey exists, a fixed benchmark is established on or near the site and every subsequent level — excavation depth, drainage fall, floor level, wall height — is set from that same datum with a rotating laser or similar instrument. Working this way means the whole site is internally consistent: a drainage run set from the same datum as the foundations will actually reach the invert level it was designed to reach.

On steeply sloping sites it is worth having the survey extend beyond the immediate footprint, capturing the access route and any area where spoil, retaining structures or drainage attenuation might need to sit. Discovering after excavation starts that the only flat ground for a soakaway is where the digger needs to stand is an avoidable problem.

Survey accuracy matters more as slope increases. A levelling error that would be trivial on flat ground can translate into a retaining wall a course too short, a drainage run with the wrong fall, or a floor level that does not clear a flood plain requirement. Where a site is genuinely complex, a survey carried out by a competent surveyor with appropriate instrumentation is worth the outlay against the cost of correcting a levels mistake once concrete is in the ground.

On any sloping site, the designer's central earthworks decision is how much material to cut from the high side and how much to place as fill on the low side. Get this balance right and spoil movement, haulage and import of hardcore are all minimised. Get it wrong and the site either generates far more spoil than it can use, or needs imported fill it did not have to.

A cut-and-fill exercise works from the survey levels and the proposed finished levels to calculate the volume of material that needs to move, and where. The aim on a difficult, remote or access-restricted site is usually to get as close to a balanced cut-and-fill as the design allows, because both importing and exporting material become disproportionately expensive when haul routes are long or narrow.

Fill is not simply moved and left. Where fill will support a building, a hardstanding or a retained structure, it has to be placed and compacted in controlled layers of engineered material, tested where required, because uncompacted or poorly placed fill settles unevenly later and can cause exactly the cracking and drainage failures a difficult site is already prone to. Fill that is only supporting landscaping has a much lower bar, but the difference between the two needs to be understood and designed for, not assumed.

Material suitability matters as much as volume. Topsoil, organic material and very wet or peaty ground are generally unsuitable as structural fill and need to be stripped separately and either used for landscaping or removed. Rock arisings from a cut can sometimes be crushed and reused as sub-base or fill, which is often the most cost-effective way to deal with rock encountered mid-slope, but this depends on the rock type and needs confirming rather than assuming.

Getting cut-and-fill wrong is one of the more expensive mistakes on a sloping plot because it is usually only discovered once excavation is under way and the true shape of the ground is exposed. A cut-and-fill calculation done properly at design stage, checked against the survey rather than an assumed uniform slope, is a modest cost against the alternative of double-handling material or importing fill that was never needed.

Slope introduces cut and fill, stepped or deeper foundations, retaining structures and drainage complexity, and it affects almost every other decision on the project. Designing along the contour rather than across it, and balancing cut against fill, has more effect on cost than almost any other single decision available to a designer.

A building set across a slope, rather than following it, tends to need a stepped foundation or a substantially deeper one on the downhill side, more retaining structure, and a more complicated drainage strategy because water naturally wants to run towards and around the lowest point of the building. Orientating the building to work with the contour, or accepting a split-level design, often produces a cheaper and more resilient result than fighting the slope to achieve a single flat platform.

Access for construction traffic and plant is affected by slope as much as design is. A gradient that a car manages comfortably can be genuinely difficult for a loaded concrete wagon or a tracked excavator on a wet day, and this needs assessing at the access route as well as within the site.

Surface water on a slope behaves predictably in one respect: it runs downhill towards the building and towards neighbouring land. Interception drainage above the building line, correctly graded ground away from the walls, and a clear route for water to reach a legitimate outfall all need designing in from the start rather than added once a damp problem appears.

Narrow lanes, weight limits, soft verges and no turning space change the method for every delivery and every pour. Tracked plant, self-delivery, smaller loads, pumping and staged deliveries are the usual answers — all of which take time and need to be reflected honestly in both programme and price.

The first practical question is whether a standard ready-mix concrete wagon can reach the pour location. If it cannot, options include a concrete pump from a wagon parked further away, smaller loads brought in by a site dumper, or supply by a smaller specialist mixer. Each has cost and timing implications, and the choice needs settling before the pour is scheduled, not on the morning of it.

Weight-restricted bridges and roads are a genuine constraint in parts of rural Cumbria, and a route that looks fine on a map can be closed to the vehicles a project needs. Checking weight limits along the actual delivery route, not just at the site entrance, avoids a lorry turning back partway through a job.

Soft verges and unmade lanes deteriorate under repeated heavy traffic, and a route that started the project intact can be rutted and impassable by the end of it if nothing is done to protect it. Temporary trackway, timing deliveries to avoid the wettest ground conditions, or upgrading the access early in the programme are all ways of keeping the route usable throughout.

Turning space, or the lack of it, affects every vehicle movement on the site. Where a lorry cannot turn, it either reverses in over a long distance — which has its own safety implications — or the site needs a temporary turning area formed before deliveries start. This is worth resolving at the planning stage rather than working it out with the first delivery driver on site.

Distance affects travel, deliveries, supply availability and utility connections. Lead times for new connections in remote areas can run to months and cost considerably more than clients expect, and this needs to be factored into the programme from the outset rather than discovered midway through.

Materials that would arrive same-day or next-day close to a town can take longer to reach a remote plot, and suppliers sometimes apply minimum order quantities or delivery surcharges for long or awkward routes. Ordering further ahead, and consolidating deliveries where practical, reduces the number of separate trips a supplier has to justify.

New utility connections — power, water, telecoms — to an isolated plot can require a long private run, easements across other people's land, and coordination with the relevant network operator. These processes are frequently the longest lead-time item on a remote self-build and are worth starting the moment the plot is secured, well before groundworks begin.

Where there is no mains water or drainage, private supply and treatment become part of the groundworks package rather than an afterthought: a borehole or private supply, and a treatment plant or septic system with a properly designed drainage field, both need siting, designing and, in most cases, permitting or registering with the appropriate authority.

Remoteness also affects labour and welfare. Journey time for a workforce, the practicality of daily deliveries of small items, and the distance to the nearest facilities all shape how a remote site is run day to day, which is covered further in the welfare chapter below.

Confirm the outfall before anything else. On high-rainfall, low-permeability ground, soakaways often will not work, and discharge to a watercourse with attenuation becomes the practical route. Percolation testing settles the question rather than leaving it to assumption, and it should be done before a drainage strategy is finalised, not after a soakaway has failed to perform.

Foul drainage on a site without a mains sewer connection needs a treatment plant or septic system sized to the property, sited with the required separation from buildings, watercourses and boundaries, and discharging either to a drainage field or, where permitted, to a watercourse. The siting of the drainage field itself can be a significant constraint on a small or steep plot, because it needs suitable, undisturbed ground with adequate fall.

Surface water strategy on a difficult site usually combines several elements: interception drainage to stop water reaching the building from higher ground, correctly graded surfaces to carry water away, and either a soakaway, an attenuation system discharging at a controlled rate to a watercourse, or a permitted connection to an existing surface water system. Which combination is appropriate depends on ground permeability, available fall and the capacity of any receiving watercourse or sewer.

All of this needs designing before excavation starts, because drainage falls, invert levels and outfall positions determine trench depths and, on a sloping site, can determine the finished floor level of the building itself. Confirming the outfall late, after foundations are already in, is one of the most expensive sequencing mistakes possible on a difficult site.

Managing water while a site is being built is a different problem from designing the permanent drainage system, and difficult sites tend to suffer on both fronts at once. An open excavation on sloping or high-rainfall ground collects water from rainfall directly and from groundwater or run-off from surrounding higher ground, and if nothing is done about it, foundations get poured in standing water and formations get disturbed before they can be inspected.

Temporary measures typically include cut-off ditches or bunds above an excavation to intercept surface run-off before it reaches the working area, sump pits within an excavation with a pump to remove water as it collects, and careful timing of excavation and pour sequences around forecast weather rather than a fixed calendar date.

Where groundwater is present rather than just surface run-off, dewatering becomes a more significant undertaking, potentially running for the duration of an excavation being open. This needs planning for in the programme and the price, because pumping equipment, standby arrangements and discharge of the pumped water all have a cost, and an uncontrolled discharge of silty water to a watercourse can itself cause a pollution problem that needs avoiding.

Protecting formations once they are exposed matters as much as removing standing water. A formation that has been softened by exposure to rain, even briefly, may no longer be suitable to build on without further work, so covering, sequencing pours quickly after excavation, or timing the work around settled weather all reduce the risk of having to re-excavate a compromised formation.

None of this is exotic engineering, but it does need planning for before the excavator arrives, because on a difficult site the working window in which ground can be safely opened and closed again is often shorter than the parties expect.

Difficult sites generate retaining walls. They are structural elements needing design, foundations and drainage behind them — not landscaping features, and not the place to save money. A retaining wall holds back a live load of soil, and often water, and its failure can undermine a building, a road or a neighbouring property.

Design input from a structural engineer is normal above a modest retained height, and is worth having even below that threshold on ground where loading is uncertain, where the wall sits close to a building, or where the consequences of failure would be serious. The engineer will specify wall type, foundation, reinforcement where relevant, and drainage provision.

Drainage behind a retaining wall is not optional. Water building up behind a wall adds hydrostatic pressure the wall was not designed to resist, so permeable backfill, a drainage layer and weep holes or a land drain at the base are standard details that need to be built as specified, not omitted to save time.

On a sloping plot, retaining walls interact directly with the cut-and-fill strategy: reducing the height and length of retaining structure needed is usually a cheaper route to a workable platform than accepting a larger retained height and building a bigger wall to suit. This is worth reviewing at design stage, when the building's position and level can still be adjusted, rather than after the wall has been priced.

Any excavation on a difficult site, whether it is a foundation trench, a cut for a retaining wall or a platform cut into a slope, has to remain stable for as long as people are working in or near it. On steep or unstable ground this is a more serious question than on a flat, stable site, and it needs a deliberate answer rather than an assumption that the sides will hold.

The options are broadly: battering the sides back to a safe angle where there is room to do so, stepping the excavation, or supporting the sides with sheeting, trench boxes or other temporary works. Which is appropriate depends on the soil type, groundwater, depth, and what loads are nearby — a wall, a road, a neighbouring foundation — that the excavation might affect.

On sites with a slope superimposed on the excavation, or with soft, saturated or made ground, the risk of a slip or collapse is higher and needs assessing specifically rather than by comparison with a flat-site trench. Where an excavation is deep, close to a structure, or in ground of uncertain stability, temporary works should be designed rather than judged on site, and this is an area where engineering input earns its cost directly in preventing injury and rework.

Excavations left open through wet weather are more prone to instability as saturation reduces the strength of the soil at the face. Minimising how long an excavation stays open, protecting exposed faces where practical, and having a plan for what happens if weather intervenes mid-excavation are all part of managing this risk on a site where the working window is already tight.

Rock, boulder clay, peat and running sand all appear in Cumbria, sometimes in the same trench. Method, plant and price should acknowledge that rather than assume uniform digging. A quotation based on ground conditions that turn out to be different is not a quotation that can be relied on, which is why an honest statement of what the price assumes matters more than the headline figure.

Rock encountered during excavation may need breaking rather than simply digging, and depending on the volume and hardness, this can mean a hydraulic breaker attachment, longer machine time, or specialist plant for larger volumes. Boulder clay can contain cobbles and boulders that slow digging and are awkward to reuse as fill. Peat is weak, compressible and generally unsuitable as a foundation-bearing material or as structural fill, and running sand can flow into an open excavation faster than it can be dug out, particularly where groundwater is present.

Where excavation reveals conditions that differ materially from what was assumed at pricing stage, the sensible response is to stop, record the position, and discuss it with the client and, where appropriate, a structural engineer, rather than building over an unsuitable formation. Photographing and recording the unexpected condition before work continues protects everyone if the change needs to be justified later, whether for building control, an engineer or a warranty provider.

Where conditions are uncertain, a ground investigation is money well spent. It converts an open-ended risk into a designed solution and gives every contractor pricing the job the same basis to work from, which also makes quotations genuinely comparable rather than each based on a different guess.

A ground investigation typically involves trial pits or boreholes at representative locations across the site, with samples taken for laboratory testing where needed to establish bearing capacity, groundwater level, and the presence of any contamination or unsuitable material. On a site known to have variable ground — glacial till giving way to alluvium, or shallow rock in one corner and peat in another — more than one or two trial locations may be justified to get a representative picture.

The output of a ground investigation feeds directly into foundation design, drainage design (particularly percolation testing for soakaways or drainage fields) and earthworks planning, so it is worth commissioning early enough that its findings can actually influence the design rather than arriving after foundation depths have already been fixed on a drawing.

Where a full investigation is not proportionate — a small extension on a plot with known, consistent ground, for instance — trial holes dug as part of the early groundworks can still be used to confirm assumptions before foundation concrete is ordered, giving at least some of the same protection at lower cost.

Rainfall, exposure and altitude compress the working window. Excavations flood, concrete struggles in frost, and masonry needs protection. Programmes should carry honest contingency rather than assuming a run of dry, mild weeks that Cumbria does not reliably provide, particularly at elevation or in open, exposed locations.

Concrete placed in cold weather gains strength more slowly, and in sustained frost, waiting for a better window is cheaper than repairing frost-damaged concrete or dealing with the consequences of a foundation that has not achieved adequate strength before being loaded. Protection measures — insulating blankets, admixtures, delaying the pour — all have a cost that needs weighing against the cost of delay.

Masonry and mortar are similarly affected: mortar that freezes before it has set loses strength permanently, and work in cold, wet or exposed conditions may need temporary protection or a pause in the programme. On exposed upland or coastal sites this is a more frequent consideration than in a sheltered valley location.

Wet weather affects earthworks directly, since saturated ground is harder to compact properly and wet clay in particular becomes unworkable and unsuitable for reuse as fill. Scheduling earthworks for drier periods where the programme allows, and having a contingency plan for when it does not, reduces the number of days lost to conditions that were foreseeable even if the exact timing was not.

In the National Park and conservation areas, planning conditions frequently specify stone type, coursing and pointing. Availability and lead times for the right stone can affect programme as much as the workmanship affects appearance, and this needs checking before a programme date is promised.

Where a condition requires stone to match an existing building or a specific local type, sourcing needs to start early: some stone types are only available from a limited number of quarries or reclamation yards, and matching an existing building exactly can take longer than sourcing a generic supply. Confirming availability before committing to a start date avoids a masonry gang standing idle waiting for material.

Mortar specification is usually part of the same condition set, particularly on traditional or listed buildings, where a lime mortar rather than a cement-based mix is often required to allow the wall to breathe and to match the flexibility of the original construction. Using the wrong mortar can cause damage to older masonry that is expensive and difficult to put right.

Where a discharge of conditions application is required before stone or mortar samples are approved, that approval needs to be built into the programme as a discrete step with the local planning authority or National Park Authority, since starting masonry work before conditions are formally discharged carries planning risk regardless of how good the workmanship is.

Prove that plant can reach the working area before the programme is set. Occasionally a site genuinely cannot take the plant a job needs, and that is better established at the site visit than halfway through, when a machine that cannot get past a gateway or across a soft field has already cost a wasted trip.

Compact tracked plant suits the restricted access typical of rural Cumbria, spreading load over a wider area than wheeled plant and fitting through gaps a larger machine cannot. Larger machines are more efficient where volumes and access allow, so matching plant size to the site rather than to habit saves both time and money — oversized plant on a tight site is as much of a problem as undersized plant on a large one.

Ground bearing capacity matters as much as gateway width. A field or unmade track that looks solid can become impassable to tracked plant once it is wet, and a site that was accessible during a dry site visit may not remain so through a wet winter. Temporary trackway or timing the heaviest plant movements for drier conditions are both ways of managing this.

Where a genuinely difficult access rules out machinery of the size a job would normally use, the alternative is usually smaller plant working longer, more careful sequencing, or in rare cases, other means of moving material such as a conveyor or, on very constrained sites, manual handling for the last stretch. All of these cost more time than standard plant on standard access, and that needs reflecting honestly in the programme.

On remote sites the distance to a licensed facility can make disposal one of the largest single costs. Reuse and regrading on site are usually the cheapest and lowest-impact route where the material is suitable, which is another reason the cut-and-fill balance discussed earlier matters so much on a difficult plot.

Where material genuinely has to leave site — because there is too much of it, or because it is unsuitable for reuse, such as contaminated ground, excess topsoil, or material with high organic content — it has to be taken to an appropriately licensed facility, with duty of care documentation completed for each load. On remote sites, the round trip for each load is longer, which multiplies both the cost and the time involved compared with a site close to a disposal facility.

Material suitable for reuse as fill or landscaping should be identified and stockpiled separately from material that cannot be reused, so that a load of good material is not accidentally sent for disposal alongside unsuitable material, or vice versa. This separation is easier to manage if it is planned before excavation starts rather than sorted out after everything is already in one pile.

Access constraints that limit plant size also limit the size of vehicle that can remove spoil, which on a very restricted site can mean smaller loads and more trips than a straightforward site would need. This should be reflected in the spoil removal allowance from the outset rather than discovered once the first attempt to bring in a full-size muck-away lorry fails.

Difficult sites sometimes need material or components lifted into position where plant cannot drive close enough, or where the item being placed — a structural steel beam, a large precast unit, a modular component — is too heavy or too large to handle by other means. On a restricted or sloping site, this needs planning as a discrete operation rather than treated as an extension of general plant use.

The first question is whether a crane, of whatever size is needed, can actually get to a position from which it can make the lift, and whether the ground it would stand on can bear the load, including the outriggers under full load rather than just the travelling weight of the machine. On sloping or soft ground this is not a given, and may require ground preparation such as compacted hardstanding or spreader mats before a lift can be attempted safely.

Where a site cannot accommodate a mobile crane at all, alternatives include a smaller crawler crane that can travel on tracks over softer or steeper ground, a long-reach machine positioned on more stable ground at a distance, or in some cases a helicopter lift for genuinely inaccessible locations, though this is a specialist and costly option reserved for sites where nothing else will work.

A lift plan, prepared by a competent person, covering the weight, the lifting radius, ground conditions, exclusion zones and any overhead hazards such as power lines, should be in place before a significant lift takes place on any site, and particularly on a difficult one where the margin for error in ground bearing or access is already reduced.

A remote site still needs welfare facilities — toilets, somewhere to eat, washing facilities and first aid provision — and providing these becomes more involved the further the site is from mains services and from suppliers who can deliver at short notice. This needs planning as part of the site setup rather than left until the workforce arrives on day one.

Where there is no mains water or power on site, welfare units typically need to be self-contained, with their own water supply, waste tank and power source such as a generator or battery system, all of which need periodic servicing that has to be planned into the logistics of a remote job — someone has to get out to empty a waste tank or refill water on a schedule that keeps pace with site use.

Material storage and site security also need more thought on a remote plot. A site a long way from habitation is more exposed to theft of tools, fuel and materials, and further from a quick response if something goes wrong, so secure storage, and sometimes on-site fuel and generator security, become a bigger part of the setup cost than on an urban site.

Telecoms coverage is worth checking early: a site with poor or no mobile signal affects everything from ordering materials to raising an alarm in an emergency, and where coverage is genuinely absent, a site-based means of communication, such as a landline installed early or a satellite device, may be worth arranging before work starts rather than discovering the gap once it matters.

Difficult sites carry more unknowns, so build float into the programme, resolve information before mobilisation, and expect the sequence to be adjusted as conditions become clear. A programme that assumes best-case ground, best-case weather and best-case access on a site that has none of those characteristics is not a realistic programme, whatever the drawing says.

The biggest single protection against programme risk on a difficult site is resolving as much uncertainty as possible before mobilisation: a ground investigation, a topographic survey, confirmed drainage outfall, a proven access route and a discharged set of planning conditions all remove a source of delay that would otherwise surface mid-project, at the point it is most expensive to deal with.

Weather-dependent activities — concrete pours, masonry, earthworks — should be sequenced with realistic contingency for the season and the site's exposure, rather than a single fixed date that assumes everything goes to plan. On an exposed or high-altitude site, that contingency needs to be more generous than on a sheltered lowland one.

Long lead-time items specific to difficult sites — matched stone, treatment plant components, utility connections, specialist plant hire, engineering approvals — should be identified and ordered or applied for as early as the design allows, because these are frequently the items that quietly determine the real start date of visible construction work, regardless of how quickly groundworks themselves could otherwise proceed.

On sloping, unstable or unusual ground, engineering input is the cheapest form of risk management available. It turns judgement calls into designed, insurable solutions, and on a difficult site there are usually more of those judgement calls than on a straightforward one.

A structural engineer's input is normal for foundation design where ground conditions are variable or uncertain, for retaining structures above a modest height, and for any temporary works supporting a significant excavation. Their design gives a documented basis for what was built, which matters for building control sign-off, for any future sale of the property, and for insurance purposes if a problem emerges later.

Beyond structural engineering, a difficult site may draw on other specialists at different stages: a surveyor for topographic and setting-out work, a drainage engineer or specialist for percolation testing and treatment plant design, a geotechnical engineer for a ground investigation, and an ecologist or arboriculturist where the site's features require it. Coordinating these inputs so each one informs the next, rather than being commissioned in isolation, is part of what makes a difficult site manageable rather than a series of separate problems.

Nothing in this report describes a reason to walk away from a difficult plot on its own. Most of the conditions covered here — slope, restricted access, remoteness, variable ground, exposure — are manageable with the right sequence of survey, design and professional input, provided they are identified and priced honestly before work starts rather than discovered expensively once it has.

The pattern that runs through every chapter here is the same: resolve information early, get the outfall, the ground and the access confirmed before committing to a design or a programme, bring in engineering and other specialist input where the site genuinely calls for it, and allow realistic contingency for weather, access and unexpected ground rather than pricing and programming for the best case.

A difficult site that has been through that process — survey done, ground investigated where warranted, drainage confirmed, access proven, planning position established — is not a difficult site anymore in any practical sense. It is simply a site with a known set of constraints that the design and the programme have already been built around. The plots that cause real problems are the ones where that work was skipped, not the ones where the ground happened to be steep or the lane happened to be narrow.

Anyone buying, designing for, or building on a challenging Cumbrian site is better served by treating the difficulty as a sequence of questions to answer in order than as a single risk to price in the round. Answered in order, most of those questions have a workable answer.