Quote of the day

A person who never made a mistake never tried anything new.

- Albert Einstein

Contact

Water-Aware Soiling Estimation and Cleaning Scheduling for Utility-Scale PV in Arid Climates

Open Energy

Water-Aware Soiling Estimation and Cleaning Scheduling for Utility-Scale PV in Arid Climates

Water-Aware Soiling Estimation and Cleaning Scheduling for Utility-Scale PV in Arid Climates

Description

Estimate photovoltaic soiling losses from images and inverter data, then optimize cleaning schedules while explicitly accounting for water scarcity.

Details

Context and Problem Statement

Dust accumulation can substantially reduce photovoltaic output in arid climates. Many studies classify clean and dirty panels without linking detection to actual energy loss or to the water cost of cleaning.

Research Question

How can soiling-related energy loss be estimated from heterogeneous signals, and which cleaning policy maximizes net benefit when economic and water costs are modeled explicitly?

Proposed Approach

Fuse visual soiling estimation with inverter-level performance-ratio deviations, use causal attribution to distinguish soiling from shading and faults, and formulate cleaning as a stochastic optimization problem under uncertainty in dust deposition and rainfall.

Expected Contribution

An end-to-end framework connecting visual detection, energy validation, and water-aware maintenance decisions.

Expected Prototype

A decision-support tool generating cleaning schedules by plant zone with expected energy gain, water consumption, and uncertainty.

Datasets

Public PV soiling images, inverter production data, and local weather and dust indicators.

Challenges

Lack of direct soiling-loss labels, confounding among degradation causes, seasonal variability, and electricity-water trade-offs.

Research Question

Which cleaning policy maximizes net photovoltaic output in arid climates when water is treated as a constrained resource rather than a fixed-cost input?

Innovation

The project links detection, energy-loss quantification, and operational scheduling under explicit water constraints.

Expected Deliverable

A validated energy-loss estimator, a water-aware cleaning scheduler, and a comparative techno-economic assessment.

Technologies

Computer Vision (CNN, Vision Transformers) Energy Production Time-Series Analysis Stochastic Optimization Causal Attribution Drone and Thermal Inspection

Required Skills

  • Computer Vision
  • Energy Production Data Analysis
  • Optimization and Operations Research
  • Photovoltaic Engineering Fundamentals

Datasets

  • Public images of soiled photovoltaic modules
  • Inverter-level production data from a partner plant
  • Local weather and dust-index data

Morocco & Africa Relevance

Morocco is expanding solar capacity in semi-arid regions where water scarcity is structural, making joint optimization of electricity yield and water use highly relevant.

References

  1. Visual Deep Learning-Based Soiling Detection on Photovoltaic Panels with Inverter-Level Energy Validation
  2. Deep Learning-Based Recognition and Classification of Soiled Photovoltaic Modules
  3. Deep Learning-Based Approach to Automated Monitoring of Defects and Soiling on Solar Panels

Project Information

Duration: 4 to 6 months
Academic Level: Engineering, Master's or PhD
Research Area: Energy

Keywords

Photovoltaic Soiling Computer Vision Maintenance Scheduling Water Scarcity Energy Yield

Evaluation

Scientific novelty
Impact
Technical feasibility
Data availability
Prototype potential
Industrial relevance

Apply for this topic

To apply for this topic, please send your CV and portfolio to:

contact@yanecode.com
Back to Research