The UK needs to do a lot more building. A lack of access to physical and digital connections is holding back the economy, the government says. Besides tackling the housing crisis, the country has to construct more roads, railways, wind farms and reservoirs to open up opportunity and drive productivity. The only problem is that Britain is notoriously inept at delivering infrastructure projects on time and within budget.

The advantage of doing things badly is that at least you get to learn from your mistakes — in theory. Updates this month have offered some illuminating insights into two of the biggest civil-engineering undertakings in the country: High-Speed Rail 2, better known as HS2, and Hinkley Point C, which will be Britain’s first new nuclear power station since 1995. Here are five lessons that can be drawn from the issues encountered by two projects with a combined bill that’s likely to exceed £100 billion ($127 billion):

Don't take budgets too seriously — especially at the start. Fixing an initial budget that was too low may have done much to feed later perceptions that HS2’s costs were spiraling out of control. The original estimate for the expanded train network was set too early and based on “very immature data,” Jon Thompson, appointed executive chair of High Speed 2 Ltd. in February last year, told the House of Commons transport committee. Numbers get more accurate and reliable as work progresses and the quality of information improves. What were viewed as cost blowouts partly reflected this process. The effect was unfortunate, undermining political support for HS2 and providing cover for cutbacks that have reduced the network to a single line between London and Birmingham that fails to fulfill most of its original purpose. To avoid this problem: Stick to a range rather than a single figure, and make sure people understand the uncertainties inherent in early-stage estimates.