Food, energy and water under climate risk

Making your own power and water pays when you share it and plan both together. Planning for a climate you cannot know costs more than bad weather.

Every number here comes from published SEAR Lab papers, listed at the bottom with the table or figure each one comes from. Where a paper has an error, the demo shows the corrected value and says so. Cite the papers, not this page.

A. An Austin neighborhood: does it pay to make your own electricity and water?

Rooftop solar, batteries, rainwater tanks and graywater recycling can cut utility bills. Choose how many homes pool their investment and what they plan for, and compare the yearly cost per home with buying everything from the utilities.

Homes that invest together

The paper's 32 homes; 320 and 3,200 are 10 and 100 copies of each.

What they plan for

Capped: each home may buy at most 500 kWh and 20,000 gallons a month from the utilities.

Yearly cost of electricity and water per home

Annual cost per home by community size and planning approach
Selected approach Other approaches Buy everything from the utilities

Share produced by the homes' own systems

Share of electricity and water produced locally, by community size
Electricity Water
What gets built
    Show the numbers as a table

    Costs are the paper's Fig. 1 bar heights (the optimized yearly cost divided by the number of homes) and the local shares its Electricity and Water Fractions, read from the scenario totals in the paper's public repository. Those fractions count stored and curtailed output, so they are not exactly the share of demand met. The household usage data behind the model are licensed and are not shown or used here.

    B. A Texas wheat farm: what is it worth to know the climate in advance?

    A 200-hectare farm short of water can pay to desalinate brackish groundwater, powered by the utility or its own solar panels, but it must build before it knows whether the next 25 years will be wet or dry. Choose what the farmer believes, then see how each climate would play out.

    The farmer believes

    The climate that arrives

    What advance knowledge adds to expected 25-year profit

    Value of knowing the climate, the weather, and of hedging, for both beliefs
    Equally likely Dry most likely

    The four climate futures: share of years at each rainfall level

    Probability of each annual rainfall level in the four climates
    5 in15 in30 in 45 in60 in a year

    25-year profit in each climate

    Profit in each climate with a plan made knowing the climate and with the plan made under the belief
    Built knowing the climate Built for the belief (hedged)
    Show the numbers as a table

    Correction note. Under the “dry most likely” belief, the paper's Table 4 gives the value of hedging (its Value of the Stochastic Solution) as $940.90. Re-running the paper's public code gives $964.89: the plan built for the average climate sits on a very flat objective, and a 0.08% difference in its capacities moves realized profit by about $20. The demo shows the published $940.90 (marked *). Every other value in Table 4 reproduces to within about $1.

    What this shows

    Both studies are about the same trade: water that is scarce or expensive can be made locally, but making it takes electricity, and the investment has to be made before you know how things will turn out.

    In the neighborhood study, a home acting alone saves a little (about $3,056 a year against $3,442 from the utilities) and mostly from water, not solar. Pooling changes the picture: community solar and wind turbines become worthwhile, and at 3,200 homes a shared water recycling facility does too, cutting the cost to about $2,418 per home, 30% below the utilities. Planning electricity and water together is the cheapest choice at every size, and only at 3,200 homes does it save more than the electricity-only and water-only savings added up. The paper also finds that home water systems running on grid power always raise emissions, because small systems use more energy than city water plants, while planning them with local solar and wind always lowers emissions.

    On the farm, what the farmer believes about the climate sets the plan, and a plan built for one climate can do badly in another. With all four climates equally likely, knowing the climate in advance is worth about $98,000 of expected 25-year profit; knowing each year's weather as well adds only about $10,000. Hedging carefully across climates gains almost nothing over simply planning for the average one. Believe a dry future is most likely, then pick “Wet”: the farm that built for drought earns about $502,000 less than one that knew the rain was coming.

    Put together, the lesson is about when and at what scale to commit. Sharing spreads the cost of lumpy investments like wind turbines and recycling plants; being right about the climate matters far more than being ready for any single year.

    Behind this demo

    Sources

    1. Jones, E.C., & Leibowicz, B.D. (2021). Co-optimization and community: Maximizing the benefits of distributed electricity and water technologies. Sustainable Cities and Society, 64, 102515. doi:10.1016/j.scs.2020.102515. Panel A: scenario definitions from Section 4; costs and Electricity and Water Fractions from Fig. 1 and installed technologies from Figs. 6–7, as committed in the paper's public repository sear-labs/water-energy-coopt-scs-2021 (results/tables/published/; archived as doi:10.5281/zenodo.22716302); emissions finding from Section 5.3.
    2. Jones, E.C., Jr., & Leibowicz, B.D. (2022). Climate risk management in agriculture using alternative electricity and water resources: A stochastic programming framework. Environment Systems and Decisions, 42(1), 117–135. doi:10.1007/s10669-021-09838-8. Panel B: Table 1 (climates), Section 3.3 (beliefs), Table 4 (values of information), Table 5 (expected profits) and Tables 6–7 (profit by climate). Reproduction check from the public repository sear-labs/fews-stochopt-esd-2022 (results/clean/value_of_information.csv; archived as doi:10.5281/zenodo.22715616).