This analysis examines every public comment submitted to that application and to three others in North Lincolnshire — 807 comments in total, 756 of which contained substantive text. It is not an analysis of whether the decisions were right. It is an analysis of what the consultation record reveals: about the nature of public concern, the structure of the communities objecting, the limitations of the consultation system itself, and the national policy framework that determined the outcome regardless of how many people objected.

The four applications

The dataset spans two local planning authorities — Wiltshire Council and North Lincolnshire Council — across approximately 18 months.

You can follow the full analysis, and code used and datasets here and here

What people said — and what they meant

The 436 substantive comments on Wiltshire PL/2024/05527 are the most analytically coherent record in the dataset. Every comment is dated, named, and individually written. The concerns they raise, in order of frequency:

• Visual impact — 127 comments

• Environment and ecology — 97 comments

• Community character — 75 comments

• Flooding and drainage — 48 comments

• Noise — 45 comments

• Traffic — 24 comments

Visual impact and environmental concern account for more than half of all objections. But the most significant finding is the distribution across time. The application attracted comment in three distinct waves, each corresponding to a revision submission by the applicant: 80 comments in August 2024, 66 in November 2024, and 170 in July 2025. Each time the applicant submitted amended plans, a substantial portion of the objecting community returned — explicitly noting that the amendment had not addressed their concern.

The retirement village

The most organised community in the dataset is the 250 residents of Wadswick Green, a retirement village directly adjacent to the proposed Donkey Field site. Sixty comments identify the commenter as a resident, and they share a common narrative: they moved there expecting a peaceful rural environment, many were unaware of the application when they arrived, and the proximity of the proposed building — as close as 200m from residential units — combined with the existing low-frequency hum from the adjacent Spring Park campus, is a central concern.

“A year ago I moved to Wadswick Green retirement village where I hoped to spend the rest of my days in this beautiful, peaceful environment. To my horror I have learned that a planning application has been submitted to build a monstrous building which will be very close to Ash Lane.”

“The Council was aware when granting planning permission for Wadswick Green of the intent that it should provide a tranquil site for elderly residents. It ensured that the bat population should be protected and cherished. This application, if granted, would be counter to the principles on which the planning of Wadswick Green was founded.”

That second comment identifies something specific: the planning permissions that created Wadswick Green included conditions requiring bat-friendly lighting and wildlife corridor protection — conditions directly relevant to the data centre application. Yet it appears that Wadswick Green residents had no formal notification right. They discovered the application through word of mouth.

The Neston & Westwells Action Group — which formed in response to the application and includes Wadswick Green residents — ultimately obtained specialist planning legal advice and submitted the 17-page rebuttal that deferred the December 2025 committee date. This was a significant act of civic organisation. It also illustrates a structural inequity: communities with the time, resources, and professional networks to obtain specialist advice can meaningfully engage with the policy framework.

The flooding argument

Drainage and flooding concerns appear in 48 substantive comments across all three waves. The framing is consistent: a drainage pre-condition from a previous planning permission on the Spring Park campus was never satisfactorily discharged, and the proposed attenuation pond (I had to look this one up : an engineered basin designed to temporarily store excess rainwater and stormwater runoff during heavy rainfall. By holding the water and releasing it slowly, it prevents local drainage networks from becoming overwhelmed, thereby reducing the risk of downstream flooding and erosion)is undersized. One commenter performs the calculation: at the site’s area of 7.3 hectares and design rainfall of 50mm/hr (Building Regulations Part H), the 700m³ attenuation pond reaches capacity in under two hours from empty.

The Drainage consultee and Highways both submitted responses. Neither addressed this calculation in the comment stream, visible to anyone following this application. Their responses are accessible via the portal’s files tab as PDF attachments — present in the record, but not surfaced alongside the public representations that raised the concern.

The final approval was subject to a condition requiring a full surface water drainage scheme before development begins — effectively acknowledging the drainage question remained open at the point of consent.

The cluster question

Seven comments reference the data centre cluster explicitly.

“I object to yet another data centre in Westwells Road. This area is not designed for a massive campus, yet the council has already allowed six to be built. The seventh will be in the middle of a small village and is wholly inappropriate in size.”

The cumulative impact of six operational or consented data centres on one road — in terms of noise, traffic, lighting, and community character — was never formally assessed. Each application is processed in isolation. It appears no statutory consultee addressed cumulative impact. The question of whether a seventh data centre on Westwells Road should be evaluated against the cumulative context of the existing six was raised only by individual residents, and is not answered in the officers’ report.

North Lincolnshire: two very different records

The Humber Tech Park application at South Killingholme (PA-2024-584) attracted 54 substantive comments, all from named, dated, locally-rooted Habrough and Killingholme residents. Their concerns are hyper-specific to place: the existing noise from the A180, the oil refineries surrounding the village, the roads that cannot handle HGV traffic. There are no shared verbatim phrases, no template structures. One commenter is 16 years old and worried about green space. Another retired from Leeds specifically to find quiet.

The application was approved in under three and a half months. The 36 conditions attached directly address the concerns raised in comments: generator testing restricted to weekday daytime hours with seven days’ written notice to noise-sensitive receptor properties; full surface water drainage scheme required before development; lighting restricted per a specified assessment. The objectors correctly identified the risks. The approval recognised them through conditions. No commenter would have known this unless they retrieved the formal decision notice after the fact.

The Elsham Tech Park application (PA-2025-643) presents a different picture. 259 of its 260 substantive comments carry no date. Many address climate and energy concerns in broadly similar terms. Some share verbatim phrases traceable to a campaign platform: the text “This scheme would require a huge amount of electricity, and has no plans for renewable energy. That means it will have a huge negative impact on the climate, and should be rejected as incompatible with our climate change targets” appears, with minor variations, in 22 separate submissions. The absence of dates means the order of submissions cannot be reconstructed. Two comments are confirmed by their own text to originate outside the UK — one signs off “Arild Andervad, Sweden,” another opens “we trust from Switzerland.” Comments from non-UK residents carry no material weight in English planning law, to the best of my understanding. (Not a planning expert)

Why every application was approved

All four applications were approved or resolved to be approved. The reason is not that the planning system ignored public concern. It is that a specific national policy framework structured the outcome before the consultation began.

NPPF paragraph 87

The committee minutes for the 10 February 2026 Wiltshire decision record that officer e directed the committee to NPPF paragraph 87, which states that planning policies and decisions must recognise the specific locational requirements of different sectors — including “clusters or networks of knowledge and data-driven, creative or high technology industries” and “new, expanded or upgraded facilities and infrastructure that are needed to support the growth of these industries (including data centres and grid connections).”

The effect at Wiltshire was counterintuitive. The existing Spring Park cluster — which residents had cited as a reason to refuse a seventh data centre — became, under National Policy Planning Framework 87, a positive material consideration. The site already constitutes the kind of cluster the policy seeks to protect and grow. The application was, on the national policy framework, structured for approval provided technical objections could be addressed by condition.

Not one of the 1,100+ objectors cited NPPF paragraph 87. The Action Group’s rebuttal focused on drainage, flooding, local plan compliance, and the Corsham Neighbourhood Plan — legitimate arguments, but ones the officer’s report had already addressed. The policy argument that determined the outcome was one residents had no straightforward way to access without specialist planning advice.

Findings

These findings are offered as hypotheses grounded in this dataset. None should be treated as an established national pattern without validation across a larger and more representative sample.

Finding 1: Aggregated comment metrics misrepresent the character of public opposition. Standard reporting of objection volumes does not capture whether objections are dated, locally-rooted, individually-written, or campaign-routed. A structured methodology for characterising consultation responses in officers’ reports does not currently exist.

Finding 2: NPPF paragraph 87 functions as a structural approval presumption for cluster sites. In practice, the explicit inclusion of data centres in NPPF 87 means that where a cluster exists, further applications are approved provided technical objections can be conditioned. The outcome is very likely influenced heavily by the policy framework, not the consultation.

Finding 3: The consultation system separates technical and public participation, likely a small design opportunity that may contribute meaningfully to discourse. Statutory consultees respond through a parallel channel not directly and visibly surfaced in the comment stream. The public cannot see whether their specific technical concerns have been addressed by the statutory body responsible for assessing them. They can if they managed to click through and view/download the files.

Finding 4: Approval is not the end of the community relationship. The Killingholme and Elsham applications are outline consents with all matters reserved. Reserved matters applications on design, scale, appearance, and landscaping must be submitted within three years. The detailed engagement with affected communities is still ahead, and this is important to note.

The deeper question — whether communities that host major data centre infrastructure should have access to mechanisms beyond standard planning consultation, given that the policy framework significantly constrains the scope of local democratic input, evidenced in these comments — is one for national policy. It was not answered in any of the four applications in this dataset. It may appear that the question is yet to be asked.

Analysis based on 807 planning consultation comments from four UK local authority planning portals. Primary sources: Wiltshire Council Strategic Planning Committee minutes, 10 February 2026 (MId=16207); PA/2024/584 formal decision notice, 28 August 2024 (North Lincolnshire Council); North Lincolnshire Council planning committee officer’s report for PA/2025/643, 11 March 2026. PA/2025/643 was resolved “minded to approve” subject to National Highways withdrawing a holding objection and S106 signing; formal permission not yet issued at time of writing. The outcome for Wiltshire PL/2023/10233 was confirmed on the planning portal. All comment quotes are drawn verbatim from the planning record.

Cycling in Central London for leisure is a bit of a minefield if you’re new to the city.

You’re nervous about leaving your ‘nice’ bike standing around, you want a great flat white at the end of the ride, and you are entirely sick of doing your 400th lap around Richmond Park or looping around Regent’s Park. One should be able to tap a route, navigate the backstreets, lock up securely, and drink a genuinely good coffee in the sun.

So I decided to solve this problem for the sample size of 1.

Sources

1. Finding chill routes I started with Transport for London (TfL) cycling data. According to the raw data, London was an absolute cycling wasteland unless you wanted to ride in circles. It turns out TfL had rebranded almost all of their “Quietways” to standard “Cycleways”. A quick fix: limit data to modern Cycleways while explicitly excluding the high-traffic commuter paths, because I have an interest in staying alive and injury-free.

2. Getting to the chill route if you are far away Live routing from TfL. Given your current location and a chosen ride’s start point, it returns journey options—if you must.

3. Finding user-preferred cafes OpenStreetMap for cafe data is great, but it doesn’t have a “hipster vibe” filter. I built a custom model that excluded chains while hunting down specific parameters: specialty coffee, on-site roasters, and most importantly, outdoor seating.

I am told that if you can’t sip your flat white while actively staring at your bike to make sure nobody touches it, it does not count as a leisure ride.

4. Being risk-informed as a London newbie Risk areas derived from Metropolitan Police neighborhood-level crime data covering the most recent 24 months. That’s roughly 25,025 thefts across ~3,608 areas. (Very interesting findings there on its own, but best saved for another day).

The Result

Some data parsing in Python, cleaning, and wire framing later, here it is.

Closing thoughts

When I overlay bike theft risk against independent coffee shops and quiet roads, I saw modern urban friction everywhere. The places with the best coffee often overlap with higher property theft density. The quietest paths leave you stranded miles away from a secure bike stand.

Next up: actually taking it out into the wild to see if the data holds up against reality. (And making sure that bike repair kit actually has a 15mm spanner).

Early feedback from the first few people who looked at it so far:

1. “Make sure the coffee lives up to expectations, maybe improve by scoring reviews?”

2. “The theft data UI is chaotic.” (Fair statement, more work definitely needed).

3. “Could this have been a Google Doc?” Probably. But not today.