Just When will there be Grid Capacity for AI? The UK Government has halted the latest contract for difference auctions on the grounds that ‘curtailment payments’ (when their isn’t grid capacity) will mean they will increase electricty bills. Lowering bills now being the priority. But that riases the issue – why not apply the same logic to Data Centres? There are major programmes to upgrade the grid – but they wont deliver result in the major bottlenecks, in the north of Scotland, til 2030 -2035 – and at a coast of around half of that of HS2. Although the term ‘auction’ is used these are in fact reverse auctions, the taxpayer pays the suppliers. Therefore is is no revenue from these to speed grid construction. If the AI boosters are right Ai will raise productivity, if so its use will raise productivity, which can be taxed to pay for upgrading the grid. Why should the taxpayer pay for energy diversification and grid upgrades when the greatest consumers of power, like AI, pay for it and thus pay for the grid upgrades we need as a society. https://andrewlainton.wordpress.com/2026/11/03/just-when-will-there-be-grid-capacity-for-ai/ https://1.gravatar.com/avatar/1af13b7db1771b3684c6167a90685dfbd7d2deda11b14381de81aa3bfeabe2c5?s=96&d=identicon&r=G #gridcapacity #datacentres #ai #rss #feeds #linksharing #qwibbit

We can’t have a total ‘let the tech bros build, build, build’ [approach],” Governments Data Centre Covenant Today the Chancellor announced a ‘Data centre covenant; priority grid access in return for reduced water consumption, local investment etc. The leak in the Financial Times suggested it was somewhat formative in development with possible undefined sustinability commitments and agreements on energy use. Notably the Chancellor only mentioned water use not energy use. Lets be clear guarenteed grid access means bumping housing development down the queue, or blackouts, it is impossible to see how 80GW of extra electricity supply can be conjured up by 2050. The sector faces a potential ‘compute commencement wall’. Already this week Microsoft withdrew support for the largest data centre in the UK by Nexus as a grid connection could not be secured till the mid 2030s. Almost all other Western Economies have introduced similar measures so they are to be welcomed. If the new SDSs are to dpo there jobs they will plan for electricity power and grid infrastructure over 25 years and make the hard choices. However it is perhaps too late. Already most AI forms have 10s of thousands of GPUs in storage as they cant build and connect data centres fast enough to plug them in. They currently cant afford to buy unless funded by the GPU manufacturers. Will this bubble burst? Already the useful life of old GPUs is being extended. But sometime in 2027 both NVidia and AMD will be introducing a new generation of GPUS which will dramatically reduced cost per unit of processing power. So my instinct is that data centre forms will load off their old stock to afford new processors, to remain competitive and get grid connections soner. This will require a dramatic write off of inventor values as they will effectively have to purchase chips twice in comparison to new entrants. Firms such as Coreweave and Oracle could be part of a nearly 1 trillon stock write down, and we could have another creit crunch. https://andrewlainton.wordpress.com/2026/10/28/we-cant-have-a-total-let-the-tech-bros-build-build-build-approach-governments-data-centre-covenant/ https://1.gravatar.com/avatar/1af13b7db1771b3684c6167a90685dfbd7d2deda11b14381de81aa3bfeabe2c5?s=96&d=identicon&r=G #centrecovenant #governmentsdata #build #financialtimes #gpus #already #westerneconomies #rss #feeds #linksharing #qwibbit

BNG to be abolished for Small Brownfield Sites Just announced in budget for sites less than 1.5Ha (in red book). Sensible as David Rudlin has written it assumes brownfield sites canbeb inflated by 10% in size. https://andrewlainton.wordpress.com/2026/10/28/bng-to-be-abolished-for-brownfield-sites/ https://1.gravatar.com/avatar/1af13b7db1771b3684c6167a90685dfbd7d2deda11b14381de81aa3bfeabe2c5?s=96&d=identicon&r=G #rss #feeds #linksharing #qwibbit

How to Stop Burnham ‘Cooling’ on New Towns The FT (paywall) – repeated in AJ (not) states: According to a report in the Financial Times, which has spoken to people close to the government’s discussions, Burnham has ‘cooled’ on the prospect of committing government funding to the programme, instead wanting housing to be delivered through urban densification and extensions led mostly by metro mayors. The newspaper did note, however, that the coming overhaul of the England-wide programme does not necessarily mean the end of all government involvement, with ‘allies of the prime minister’ saying that Burnham has not yet ruled out retaining some centrally backed schemes. The prime minister is apparently not keen on spending money from Westminster coffers on large-scale developments at the same time as attempting to limit the scope of tax rises included in the upcoming budget, to be announced next week. There are two aspects to the story – first re devolution the second regarding public spending. The Starmer led New Town Programme was undoutedly a National Programme, it did not arise from any form of strategic planning and the SEA is simply compares the national programme on and off scenarios. As such it does not chime with Burnhams push for devolution to strategic mayors and development of SDSs. This creates problems however, some proposals are well suited to for example Mayoral Development Corporations – such as Brabazon, however whole regions such as Liverpool and the West Midlands were excluded from the programme because of a bizarre treasury GIS model which only prioritised the highest performing metros and regions such as Bristol, Leeds and Manchester (although in the latter case bizarrely ignoring its own New Town proposals – a fact likley remembered by Burnham). Also many new New owns will be proposed SDSs will proposed New Towns whilst some in Reform and Conservative dominated areas are likely to defy government and protect all Green Belt and likely propose none. This will create delay. Also some areas such as Milton Keynes will have no Metro Mayors just toothless Regional Planning Committees. The second problem is that New Towns will require up front government investment (via eith central funding directly, or indirectly to mayors) to purchase land (at existing use value), for initial infrastructure and for any affordable housing component (the last should be discarded as this is required for major new housing anywhere). From the 2026 budget the government will use public sector net liabilities rather than public net debt as its preferred debt measure. This could unleash extra borrowing when it acquires an asset with long term returns (such as land) but this budget the fiscal priority may be reducing the level of debt to calm the markets. The government would be unwise to not drop the New Towns programme as that would be seen as emblamatic and a defeat in areas such as Enfield. It should rather prioritse areas areas without mayors (such as Tempsford and Milton Keynes) and back publicly strategic mayors to creates MDCs. Other RPCs and Mayoral Strategic Authorities however need some assurance in setting up MDCs and how much money to spend. This also applies to major transport investments that have slipped out of the narrow 5 year national rail and road programmes. I suggest a three tier split 5 year funded programmes A 10-15 year National Investment programme. Comprised of schemes evaluated and prioritised through SDSs where government would pay a proportion and DCs (whether Mayoral or not) would pay the rest recoupted in the long term through sale of assets. This would not count as spending until they roll into 1 above. Rather it would be taken from an assigned % of GDP (like the National Infrastructure 3% rule) with the assumption of a minimum discounted return over 25 years (such as 25%). This would include borrowing to fund utilities investment (such as the grid and water). For 15-25 tear investments simply safeguard the land with the SDS soundness test only being that they have a reasonable chance of recovery in the long term. This would resolve the glaring gap in SDS planning at the moment, how to ensure the long term infrastructure to deliver investment unlockling housing is deliverable and not simply delayed beyond 5 years every spending review or capital spending crunch. Then the problem is how not to overstress public spending. Here the model would be the Hong Kong R+P model The key is a business model called “Rail plus Property” (R+P). For new rail lines, the government provides MTR with land “development rights” at stations or depots along the route. To convert these development rights to land, MTR pays the government a land premium based on the land’s market value without the railway. MTR then builds the new rail line and partners with private developers to build properties. The choice of private developer is made through a competitive tender process. MTR receives a share of the profits that developers make from these properties; this share could be a percentage of total development profits, a fixed lump sum, or a portion of commercial properties built on the site. By capturing part of the value of the land and property around railway lines, MTR generates funds for new projects as well as for operations and maintenance. That is why it does not need government subsidies or loans. This could be implemented by effectively auctioning shares in MDCs/DCs, There would need to be some tweaks to the New Towns Act to enable this (remembering that there masterplans can be used as LDO like planning consents) and national policy stating that for new towns only MDCs/DC can grant consent to ensure the uplift. The only issue would be infrastructure that is general – like a STP, covering a wide area rather than a specific areas around transit, this would require wider reform around capturing land value uplift on receipt of planning permission. CIL is the means to do so must needs to be split into a strategic area based charge (like London) and the small beer local charge. The beauty of this is it provides a generic approach to long term capital funding. The Treasury would resist this but the Treasury is no longer in charge of economic growth or infrastructure, that is number 10 north. Crucially it gives SDss reral bit rather than the risk of them being windy long term wish documents. The Budget should committee central funding to a priority three + milton keynes with the assumption that most we will recovered in the long terms and announce a review into a new system of funding for DCs/MDCs. There are many international models for such infras led land value capture, it should not be difficult or take long. https://andrewlainton.wordpress.com/2026/10/21/how-to-stop-burnham-cooling-on-new-towns/ https://1.gravatar.com/avatar/1af13b7db1771b3684c6167a90685dfbd7d2deda11b14381de81aa3bfeabe2c5?s=96&d=identicon&r=G #rss #feeds #linksharing #qwibbit