we-solve (beta)

Econ-487 · economics · Economics

Long-run sustainability of electricity minigrids

LLM status Open unverified

Statement

ECON-067 | R | Source-backed agenda; exact formalization is editorial

Definitions and formal target

A minigrid serves households and firms with demand $q_{it}(p_t,r_t)$, where $p_t$ is tariff and $r_t$ reliability. Capital depreciates and requires maintenance $m_t$. Revenue is $p_t\sum_iq_{it}$; operating and replacement costs are $C_t(m_t,q_t)$.

For an implementation regime $z$, define discounted operating surplus \[ \Pi(z)=\mathbb E\sum_{t=0}^{T}\delta^t [p_tq_t-C_t(m_t,q_t)]-K_0. \] Determine which tariff, maintenance, and governance regimes achieve reliable service and affordable access with a declared subsidy path, including after initial donor support ends.

Clarifications and required assumptions

Specify whether $T$ is finite or infinite, and include capital replacement, battery renewal, maintenance and terminal liabilities. A positive finite-horizon surplus does not establish permanent viability. For an infinite horizon require discounting, finite expected cash flows and solvency; also state liquidity constraints at intermediate dates. Add the declared subsidy path to revenues and record its fiscal cost separately. Define reliability, tariff collection, demand, connection/access obligations, affordability and governance. Subsidy-free profitability, viability with public support and social value are different claims.

Variants and changes of scope

  • Private/community/public ownership changes finance and governance, not just a tariff.
  • Operating balance differs from full life-cycle balance including replacement and capital recovery.
  • Donor exit, main-grid arrival and changing demand require explicit transition scenarios.

Audit conclusion and source relationship

[S1], correctly authored by Robyn Meeks and Meera Mahadevan, and [S2] support minigrid sustainability. The compilation's electricity heading has an inconsistent author attribution; this catalogue uses the primary review's authors.

References and exact source roles

[S1] Robyn Meeks and Meera Mahadevan (2025). Electricity Infrastructure. VoxDevLit 15(1), May 2025.

Verified locator / source role: Primary publisher page checked for review identity, authors, version and background. The specific published agenda is corroborated in [S2]; the exact formal target is editorial.

https://voxdev.org/voxdevlit/electricity-infrastructure

[S2] Oliver Hanney / VoxDev (living compilation). Evidence gaps: Key unanswered questions in development economics. Accessed 8 October 2026. The page currently covers 20 topics; the "16-topics URL is a historical slug.

Verified locator / source role: Topic heading: Electricity Infrastructure. Supports the broad research agenda; this is not a verbatim quotation or an attribution of the displayed mathematical target.

https://voxdev.org/topic/evidence-gaps-key-unanswered-questions-across-16-topics-development-economics

Common conventions: Additional-source status and mathematical conventions

Of the 100 supplied ECON entries, 65 distinct problems appear here; 32 close matches refine earlier entries, and three duplicate questions map to existing entries. Appendix app:audited-crosswalk lists every destination. The attachment's P/R/S/U distinctions and source audits are retained. Its precise mathematical formalizations are editorial unless the individual audit says otherwise. Candidate subsidy targets ECON-004--006 retain their U flags; the resolved approval-core question ECON-007 updates OP-0202 above. The following supplied conventions govern both this part and the attributed ECON refinements elsewhere in the catalogue, subject to local restrictions.

Parameterized questions. A problem family lists inputs that must be fixed before an instance is evaluated: population, horizon, policies, information, data law, model class and welfare weights. These are required choices, not quantities the citations are assumed to specify. Undefined applications should not be treated as identified estimands.

Indices and integrability. Sets of agents, firms and regions are finite unless a population measure is explicitly used. $i,j$ index units, $r$ regions, $t$ dates and $h$ horizons. $\mathbb E$ and $\Pr$ refer to the declared population and law; $\mathbf1\{A\}$ is an indicator. Every displayed expectation and discounted sum needs existence in the stated domain. Infinite horizons use a discount in $(0,1)$ and a convergence condition; finite horizons may use discount one.

Optimization and existence. A displayed $\max$, $\min$, or $\arg\max$ is conditional on attainment. For finite feasible menus this is immediate; general spaces need continuity and compactness/coercivity or another existence argument. Without attainment, the target is the supremum/infimum and arbitrarily near-optimal feasible choices. $\inf\varnothing=+\infty$. Empty feasibility and an unidentified target are different issues.

Potential outcomes. $Y_i(z)$ is the outcome under a specified intervention, not the observed conditional mean among people choosing $z$. In binary comparisons $\Delta Y_i=Y_i(1)-Y_i(0)$. Specify assignment, treatment versions, compliance, information and observation horizon. Interference is allowed under a full policy vector; compressed exposure notation needs a justified mapping. No interference, consistency and overlap are substantive assumptions, not automatic consequences of notation.

Populations and observation. Fix baseline eligibility and population weights. Conditioning on post-policy employment, survival, relocation or program uptake can change the population being compared. Death is not ordinary loss to follow-up; outcomes truncated by death need a composite convention, joint survival target, or explicitly defined survivor stratum. Fertility-changing interventions need a population functional rather than an unexplained fixed descendant cohort. Measurement and missingness models must be supplied.

Identification. A structural model $m\in\mathcal M$ implies observable law $P_m$ and target $\theta(m)$. Identification means \[ P_{m_1}=P_{m_2}\ \Longrightarrow\ \theta(m_1)=\theta(m_2) \quad\text{for all }m_1,m_2\in\mathcal M . \] Without identification, the target is \[ \Theta(P;\mathcal M)=\{\theta(m):m\in\mathcal M,\ P_m=P\}. \] Writing $\theta(P)$ before establishing identification can hide incompatible latent models. Confidence sets additionally need a sampling law, confidence level, asymptotic sequence and uniformity class. Point identification, coverage and informative precision are separate properties.

Mediation. Total effects of randomized assignment are distinct from cross-world natural indirect effects. Belief/effort-channel effects require a meaningful mediator intervention and additional measurement, exclusion or mediation assumptions. Randomization of the initial policy alone does not supply those assumptions. Report bounds or interventional alternatives when the stronger target is unsupported.

Equilibrium. A counterfactual $W(z)$, $p(z)$, or $\mu_t^z$ requires individual optimization, feasibility, government/resource budgets, market clearing and state-transition laws. State equilibrium existence and selection, or report ranges over compatible equilibria. Initial-state integration includes subsequent endogenous transitions; current-state integration must use the policy-dependent distribution. Reduced-form invariance after policy is not assumed.

Welfare accounts. Scores, utility, health, dollars and ecological quantities require explicit conversion before addition. Fees, taxes, subsidies, wages and extortion payments are transfers between parties; distributional weights can make them welfare relevant, but their face value is not automatically a resource loss. State ownership, geographic boundaries, foreign-income weights and fiscal finance. Do not add profits to household welfare already including dividends, or count land capitalization and the underlying benefit twice. A benefit-cost expression must distinguish gross benefits from net welfare and subtract every real cost exactly once.

Derivatives and risk. Log derivatives require positive arguments; local responses require admissible perturbations and specified equilibrium branches. Differentiation under expectations needs regularity/domination, and boundaries may allow only one-sided derivatives. Risk uses a specified law; ambiguity uses a declared nonempty law set on a common history space. Adaptive policies must be nonanticipative. A robust criterion is an editorial choice unless attributed explicitly.

Scope overlaps. Allocation, collective choice, contracts, household bargaining and coordinated policy overlap economics with optimization and game theory. They are retained because they appeared in the original catalogue; no claim of a strictly game-theory-free selection is made.

Statement licence: CC-BY-4.0 · Source file

Contributions 1

Econ-487-A1 · Exploration · LLM-generated · gpt 6 astra ultra ·

GPT 6 Astra Ultra attempt on Econ-487

Minigrid sustainability after donor exit: reserve requirements, renewal and reliable service

Problem ID: L94-483. Status: unresolved; restricted research attempt.

Life-cycle value and the ability to pay

The target asks which tariff, maintenance and governance regimes can maintain affordable, reliable electricity after initial donor support. Discounted operating value alone does not answer this. We characterize minimum cash reserves for permanent service in a finite-state operating model, and derive exact renewal-cycle conditions. These results include replacement obligations and intermediate liquidity; demand, collection and technology parameters still require empirical identification.

Meeks and Mahadevan's review [MM], publisher summary inspected on 9 October 2026, discusses cost recovery, reliability, uncertain demand and minigrid viability when grid service becomes available. Its identity is VoxDevLit 15(1), May 2025. Our viability recursion is a mathematical specialization, not a theorem attributed to that review.

Time is discrete and infinite. Cash is measured in constant currency with zero real interest on reserves and no borrowing. Let $x$ belong to a finite state set $\mathcal X$. It records battery age, equipment condition, customer demand class, governance/collection quality and grid-arrival status. At each date, after observing $x$ but before observing next period's shock, the operator chooses $u\in\mathcal U(x)$, a finite nonempty set. An action specifies tariff, maintenance and connection obligations. Actions that violate a stated affordability cap, access obligation or minimum reliability are excluded from $\mathcal U(x)$.

Possible next states form a nonempty finite set $\mathcal N(x,u)$. The period's net cash is \[ g(x,u,y)=p(x,u)\,q(x,u,y)\,\chi(x,u,y) +\sigma(x,u,y)-C(x,u,y), \tag{1} \] where $\chi\in[0,1]$ is tariff collection, $\sigma$ is the declared subsidy, and $C$ includes all operating, maintenance and replacement cash payments. Investment resources are purchased at those cash costs; any transfer component must be separated for social welfare. Donor exit is represented by states in which $\sigma=0$. Known finite subsidies can be encoded by a time state before exit. Cash $b$ becomes $b+g$ at the end of the period and must never be negative. No expenditure is due earlier within a period; if it is, divide that period into payment subdates.

A planned grid-arrival closure is permissible only if the specified service obligation allows it. Such an action pays decommissioning and terminal liabilities, enters an absorbing state with $g=0$, and cannot silently discharge a continuing reliability obligation.

An exact permanent-service reserve recursion

Define the reserves required for zero further periods by $r_0(x)=0$ and recursively \[ r_{n+1}(x)=\min_{u\in\mathcal U(x)} \max\left\{0,\max_{y\in\mathcal N(x,u)} [r_n(y)-g(x,u,y)]\right\}. \tag{2} \] The criterion is pathwise solvency for every permitted sequence of shocks, not positive expected value or a high survival probability. This is useful for a guarantee contract; less conservative probabilistic criteria are different research questions.
Theorem (Minimum reserves for indefinite reliable operation): For each $n$, $r_n(x)$ is the smallest initial cash allowing a nonanticipative policy to meet all service and cash constraints for $n$ periods. The sequence is nondecreasing. Put $r_\infty(x)=\sup_n r_n(x)\in[0,\infty]$. A finite reserve $b$ permits indefinite operation from state $x$ if and only if $b\geq r_\infty(x)$. Whenever $r_\infty(x)<\infty$, an action can be selected depending only on the state such that all its successor states have finite requirements and \[ r_\infty(y)\leq r_\infty(x)+g(x,u,y) \quad\text{for every permitted successor }y. \tag{3} \] Selecting such actions at every finite-requirement state yields an indefinitely feasible stationary policy.
Proof: For one additional period and fixed action $u$, feasibility requires $b\geq0$ and $b+g(x,u,y)\geq r_n(y)$ for every successor, which is exactly the inner maximum in (2). Finite action sets ensure that its minimum is attained. Backward induction proves the finite-horizon claim. The recursion is order-preserving and $r_1\geq r_0$, so induction gives monotonicity. Every infinite feasible policy is feasible for every finite horizon, proving necessity of $b\geq r_\infty(x)$.

Suppose $r_\infty(x)<\infty$. Choose a minimizing action at each horizon. Some action $u$ occurs along an infinite subsequence, because the action set is finite. On this subsequence (2) implies $r_n(y)-g(x,u,y)\leq r_{n+1}(x)\leq r_\infty(x)$ for each successor. Taking its unbounded-horizon limit yields (3), so every successor has a finite requirement. Choose one such action for each finite-requirement state. Starting with $b\geq r_\infty(x)$, (3) implies that next cash is at least the next state's requirement and hence nonnegative. Induction establishes this invariant for every finite date and every permitted shock path, which is the infinite-horizon requirement. This proves sufficiency without exchanging an infinite minimum and limit. ∎

Equation (2) is a verifiable engineering and finance certificate. It can exclude a low tariff even when that tariff generates high usage, because replacement reserves cannot accumulate. It can favor a maintenance program if the added near-term cost reduces future replacement or outage obligations. Actions here already impose reliability; the theorem does not prove a physical reliability model from accounting alone.

Closed-form renewal accounting

Consider a deterministic plant that starts with a new battery purchased as part of initial real cost $K_0$. At the end of each period it earns collected operating margin $g=pq\chi-c$, with fixed reliable demand $q>0$, collection $\chi>0$ and recurring resource cost $c$. At dates $L,2L,\ldots$ it must pay $R>0$ to replace the battery. Replacement cost is not included in $c$ or $K_0$ beyond the initial battery. Demand, tariff and reliability are constant and satisfy the declared access obligations.
Proposition (Cash feasibility and full capital recovery): With no subsequent subsidy and no borrowing, permanent operation is possible with some finite cash reserve if and only if $Lg\geq R$. When that condition holds the minimum reserve is zero under the stated end-of-period payment convention. With discount $\delta\in(0,1)$, full discounted project surplus is \[ \Pi=-K_0+\frac{g\delta}{1-\delta} -\frac{R\delta^L}{1-\delta^L}. \tag{4} \] Therefore subsidy-free operation with nonnegative project surplus and an affordability cap $p\leq\bar p$ is possible precisely when a tariff in the admissible interval satisfies \[ pq\chi-c\ \geq\ \max\left\{\frac RL, \frac{1-\delta}{\delta} \left(K_0+\frac{R\delta^L}{1-\delta^L}\right)\right\}.\tag{5} \] Here $q,\chi,c$ are held fixed across the tariff interval; endogenous demand requires solving (5) using its actual tariff-dependent margin instead.
Proof: At the end of date $t=kL+j$, $0\leq j<L$, cumulative cash from operations and replacements is $k(Lg-R)+jg$. If $Lg<R$, it tends to negative infinity along $j=0$, so every finite initial reserve fails eventually. If $Lg\geq R$, then $g>0$ and every prefix is nonnegative, giving feasibility from zero reserve. Discounted receipts and replacements are geometric series, giving (4). Rearranging $\Pi\geq0$ gives the second lower bound in (5), while permanent operation gives the first. Both conditions are necessary and sufficient within the constant-margin model. ∎

For an arbitrary deterministic periodic cash sequence $g_1,\ldots,g_L$ including replacements, the same argument gives a useful extension: finite-reserve feasibility holds exactly when $\sum_{j=1}^L g_j\geq0$, and its minimum reserve is \[ b^*=\max\left\{0,\max_{1\leq k\leq L} -\sum_{j=1}^k g_j\right\}. \tag{6} \] Indeed a negative cycle total eventually defeats every reserve; a nonnegative cycle total makes the worst prefix occur in the first cycle. Thus upfront maintenance can require reserves even when (5)'s simple end-of-period convention does not.

Why positive discounted value can coexist with certain failure

Take $L=2$, $g=1$, $R=3$, $K_0=1/100$ and $\delta=1/10$. Equation (4) gives \[ \Pi=-1/100+1/9-1/33>0, \] but each renewal cycle loses one unit of cash. Every finite reserve runs out. This exact counterexample relies on no statistical estimation. Discounting future losses heavily can create positive NPV while a no-borrowing permanent-service requirement is impossible.

A stochastic version is equally restrictive for guarantees. Suppose one state has a permitted self-loop producing $g=-d<0$ under every available action. Repeating that self-loop for $n$ dates requires at least $nd$ cash, hence (2) implies $r_\infty=\infty$ there. If this adverse loop has positive conditional probability each time, any finite reserve faces a positive probability of failure, although expected period margin may be positive. This does not say eventual failure has probability one; that stronger assertion would require additional drift and recurrence facts.

Economic use and unresolved questions

For a specified finite action menu, the reserve theorem completely solves the robust financing subproblem. One may compare governance regimes by changing measured collection and transition primitives, not by assigning ownership a favorable coefficient. An initial donor grant can cover $K_0$ and reserve $b^*$ but cannot cure a permanently negative cycle margin. Ongoing subsidy changes (1) and must be recorded as a public cash outlay; it is a transfer in a combined equal-weight welfare account, while battery, fuel, maintenance and administration are real costs. Household benefits are gross utility less these real costs, with tariff payments and operator receipts canceled once. This distinguishes commercial return, supported viability and social desirability.

The full target still requires credible tariff-dependent demand, collection behavior, maintenance effectiveness, affordability thresholds, transition probabilities and grid-arrival scenarios. A next empirical step is to estimate a feasible action menu with confidence bounds on those primitives, then report both reserve guarantees and a separately specified probability-of-failure frontier. Finite-horizon operating profit without those obligations cannot certify permanent service.

Completion Estimate

55%

This is a post hoc, heuristic estimate for the full catalogue target. The attempt characterizes permanent-service cash reserves and derives exact renewal-cycle solvency, replacement, capital-recovery, and affordability conditions in restricted operating models. Full minigrid regime comparisons still require tariff-dependent demand, governance and maintenance effectiveness, grid-arrival transitions, and empirically validated reliability and affordability primitives.

References

Go to contribution →