Gold Standard ICS Methodology: IoT Monitoring Requirements and What They Mean for East African Cookstove Developers

From January 2024, Gold Standard mandated that all new improved cookstove (ICS) programmes registering under GS4GG with more than 50,000 stove distributions must implement IoT-enabled usage monitoring. The mandate was not a surprise, Gold Standard signalled the direction of travel in its 2022 MRV review, but its implementation timeline has caught several East African programme developers unprepared. Projects that designed their MRV systems around kitchen performance tests (KPTs) and household surveys are now facing a fundamental rearchitecting of their monitoring approach.

This methodology review examines what the IoT monitoring mandate actually requires, how it changes the economics of East African cookstove programmes, what technology options are available and at what cost, and how programme developers should design their MRV systems to be both compliant and commercially viable.

What the Gold Standard ICS Methodology Actually Requires

The Gold Standard Improved Cookstoves Methodology (GS ICS), the successor to the older GS-TPDDTEC framework for thermal energy displacement, requires project developers to demonstrate that the stoves distributed are actually being used to displace fossil fuel or non-renewable biomass cooking. The methodology calculates emission reductions as: (baseline fuel consumption × emission factor) minus (project fuel consumption × emission factor), adjusted for the non-renewable biomass fraction (fNRB) of the cooking fuel and the fraction of the population practising open-fire cooking.

The critical parameter is usage, specifically, the fraction of households that are actively using the distributed stove as their primary cooking device at the time of monitoring, and the frequency and duration of use. Historically, GS ICS methodology allowed this to be estimated through periodic KPT and controlled cooking tests (CCTs), supplemented by household surveys. KPTs measure stove thermal efficiency in controlled conditions; household surveys estimate actual usage rates and fuel consumption. The combination produced a conservative but auditable usage estimate.

The Problem with KPT and Survey-Based MRV

The fundamental weakness of KPT and survey-based MRV is its point-in-time character. A household survey conducted by a field team captures usage at a single moment and projects it across the monitoring period. It cannot detect seasonal usage variation, a critical issue in East Africa, where cooking fuel and stove usage patterns shift significantly between the dry and wet season. It cannot detect stove abandonment occurring between monitoring visits. And it is vulnerable to reporting bias, households told that their continued use determines whether they receive project support may overreport usage to researchers.

Gold Standard's post-2018 review of large-scale ICS programmes found systematic over-crediting linked to these biases. Usage fractions reported in programme monitoring reports, typically 75–90% at 12-month follow-up, were inconsistent with the stove abandonment rates documented in independent academic studies of the same programme types, which found 30–50% abandonment within 18 months in East African contexts. The result was the IoT monitoring mandate.

What IoT Monitoring Requires in Practice

IoT-enabled usage monitoring means installing a sensor on or near each distributed stove that records cooking events, typically defined as a heat event exceeding a threshold temperature (commonly 45°C or 60°C for a defined duration), and transmits this data to a central server for aggregation and analysis. The monitoring system must provide: near-real-time usage data at the household level, a data completeness rate of ≥85% of registered devices across the monitoring period, tamper detection, and data export in a format verifiable by the VVB.

The sensor must be installed at or during the time of stove distribution, not retrofitted to existing stoves, and must remain attached and functional throughout the crediting period. For a 100,000-stove programme distributed over three years, this means managing a live IoT sensor network across potentially 50–100 geographic distribution zones, with cellular or LPWAN data connectivity, device replacement protocols for failed units, and a data management platform capable of ingesting, cleaning and archiving several million data points per day.

The Economics of IoT MRV for East African Programmes

The IoT mandate materially changes the project economics for smaller and medium-scale ICS programmes. For a 50,000-stove programme, right at the threshold, sensor procurement at $12 per unit adds $600,000 to the capital cost. Data management platform and connectivity costs for a three-year monitoring period add approximately $180,000 to $350,000 depending on geographic distribution and network topology. Field maintenance, replacing failed sensors, sim cards and connectivity, adds $60,000 to $150,000 over the same period. Total additional MRV cost: $840,000 to $1.1 million for a 50,000-stove programme.

At $8 to $12 per tonne, the pricing range for unattributed GS ICS credits in early 2026, a 50,000-stove programme typically generates 15,000 to 25,000 tCO₂e per year, or $120,000 to $300,000 in annual credit revenue. The additional IoT MRV cost represents 30–70% of annual credit revenue for the first three years. For programmes where the stove retail price cross-subsidises the carbon credit revenue, this is not viable without either a price premium on the GS ICS credits (increasingly available from buyers who specifically require IoT-verified usage data) or a larger programme scale that distributes the fixed IoT infrastructure cost across more credits.

What This Means for Methodology Selection

For programmes below 50,000 stoves, the Gold Standard ICS methodology remains available without the IoT mandate, KPT and survey-based MRV is still permitted. For these smaller programmes, GS ICS remains the strongest certification pathway from a co-benefit and SDG premium perspective. For larger programmes, the IoT mandate is a structural feature that must be incorporated into the project design from inception.

Verra's AMS-II.G v11, the recognised carbon standard frameworks small-scale methodology for improved cookstoves, now widely used under VCS, introduced its own IoT monitoring recommendation in 2023, but has not yet made it mandatory for programmes below 200,000 stoves. This creates a near-term window where large-scale ICS programmes seeking to avoid the IoT cost burden can register under VCS/AMS-II.G with survey-based MRV, at the cost of the GS SDG premium and potentially lower buyer price points. Supacare's methodology review for any ICS programme above 30,000 stoves begins with an explicit analysis of this standard selection trade-off.

VVB Readiness Under IoT-Monitored MRV

VVBs conducting Gold Standard ICS verifications under the IoT mandate are developing their audit protocols in real time. Bureau Veritas and DNV, the two most active GS ICS VVBs in East Africa, have published updated audit guidance confirming that they will require: API access to the IoT platform data during desk review; a statistical sample of device cross-checks during field verification (targeting 3–5% of registered devices for physical inspection and data consistency check); proof of GSMA-compliant data privacy consent from stove recipients; and a data completeness analysis demonstrating the ≥85% threshold.

Developers entering their first IoT-monitored verification should budget for a longer desk review phase, VVBs currently estimate 8–12 weeks for IoT data review versus 4–6 weeks for survey-based programmes, and should conduct a pre-verification internal audit of data completeness before submitting the monitoring report. Programmes that submit monitoring reports with data completeness below 85% will receive a CAR requiring either data remediation or a proportional deduction in the credit claim, the latter being the more likely outcome and the more commercially damaging.