Govt Has No Idea How Much Solar Power Pakistan Is Actually Making: Study

Pakistan’s electricity sector is facing a big data gap in rooftop solar adoption. The actual installed solar capacity is nearly 2x higher than officially documented capacity.

A joint report by HeraldX and Renewables First titled “The Uncounted Gigawatts: Distributed Solar Mapping and Grid Impact Analysis in Pakistan” combines artificial intelligence with power system modelling to assess the scale and impact of distributed solar generation in the country.

The study finds that Pakistan’s grid is already handling nearly twice the rooftop solar capacity than what is reflected in official datasets. Nationally, independent estimates place distributed solar capacity at around 38GW, compared to approximately recorded in official figures.

Lahore Shows Big Gaps in Reporting

In one heavily solar-penetrated grid area of Lahore, the AI model identified 177MW of installed rooftop solar capacity, compared with an official estimate of 95MW.

Researchers say the missing capacity largely consists of installations outside or partially outside net-metering records, making them invisible to planners despite directly affecting grid behaviour.

Grid impact and technical findings

Using PSS/E power system simulation software, the study evaluated grid performance under official versus actual solar penetration levels.

Key findings include:

• Reverse power flows nearly doubled under actual solar conditions
• Transformer loading increased from **34% to 76%** during daytime operation
• Voltage levels exceeded Grid Code 2023 compliance thresholds in both scenarios

The report warns that Pakistan’s grid is already operating under conditions not fully captured by existing planning data.

A simulated 10MW battery energy storage system improved voltage stability and reduced reverse power flows but caused power factor deterioration under lower solar penetration scenarios, indicating that storage alone is not a complete solution.

Locally built AI model

The AI system developed by HeraldX was trained specifically on Pakistani rooftop imagery, unlike generic models designed for international markets.

According to Syeda Mahnoor Tauqeer, Senior AI/ML Engineer at HeraldX, Pakistan’s energy transition is being driven by households and businesses, and requires locally developed tools to accurately measure and manage its scale.

Call for urgent data-driven planning

The study argues that the mismatch between actual and recorded solar capacity is widening and creating risks for distribution companies and system operators.

Abdul Rehman, Manager Power Markets at Renewables First, said grid planning can no longer rely on outdated datasets and stressed the need for immediate adoption of modern monitoring and modelling tools.