Every once in a while someone will publish a roadmap for running the world (or a state) on 100 percent renewable energy by some date, say 2030 or 2050. The latest considers what it would take to run New York State with sources such as wind and solar. The resulting headlines look great, and a lot of people walk away with the general impression that, if we wanted to, we could easily drop fossil fuels and nuclear power.

But delve into these roadmaps, and you’ll often find jaw-dropping numbers of solar panels and wind turbines, radical changes to existing infrastructure, and amazing assumptions about our ability to cut energy use that make switching to renewable energy seem more daunting.

This post isn’t the place for a detailed analysis of these reports, although that should follow in a future story. But for now, it’s worth pointing out that in evaluating these sorts of studies, it seems like we should ask a few questions.

First, can we really reduce energy consumption as much as these studies often claim? In order to produce enough energy from wind and solar and other renewable resources, these studies typically posit a scenario in which we will actually use less energy in the future than we do now because of improvements in efficiency. The New York State one estimates a 37 percent drop in energy demand. Is this really possible, not just in theory (we certainly waste a lot of energy) but also in practice? The New York study proposes using district heating as a way of storing energy from wind turbines, but installing the underground steam pipes and other equipment needed for such systems is a costly and likely very slow process—especially if the system needs to be retrofitted to existing cities. If we look at the situation worldwide, reducing energy demand will be particularly difficult as poor countries try to bring basic energy services to their people, which could actually double energy demand.

The other big question is, how much will the switch to renewables cost compared to the cost of continuing to use fossil fuels? When talking about the cost of switching, studies sometimes include estimates of externalities associated with fossil fuels, which is important and necessary. The New York State study estimates that air pollution plus the impacts of climate change actually double the cost of power from coal. Yet no one really knows how much damage will come from global warming, so the estimates could be inaccurate. If we do switch aggressively to renewable energy, presumably the cost for fossil fuels will drop. By how much? What does that do to our calculations about the true cost of fossil fuels, including externalities?

If the policy that drives the switch to renewables depends on adding a price for carbon emissions, falling fossil fuel prices could require an ever higher carbon price to prompt a switch. If our policy is a mandate that limits the use of fossil fuels, at what point does prohibiting fossil fuels hurt poor people who could benefit from a cheap source of energy more than they’re hurt by externalities?

Another key question about costs has to do with financing. When we’re talking about renewable energy, we’re talking essentially about paying for all of the power we’ll use over the lifetime of a solar panel upfront. The cost savings from efficiency measures also require an upfront investment. The cost and availability of financing will have a big impact on the cost per kilowatt hour of renewable energy, or whether battery-powered vehicles pay for themselves in fuel savings.

And a big unknown is just how much it will cost to integrate huge amounts of intermittent renewable sources of energy to create reliable power. The New York study gestures to this problem, but the methods proposed are untested on a large scale, and the challenge will vary considerably depend on renewable resources in a given region. In some parts of the world, doldrums set in for entire seasons, making wind power a terrible option.

I suspect the answers to these questions will suggest that switching to renewables—especially if low-carbon nuclear power is left out of the mix—is more difficult than it might seem at first, but I don’t know for sure. It’s worth a close look.