Learnings

Outcomes

Key outcomes of the project are summarised in Table 2 in the supporting document.

Recommendations for further work
Next steps for this work is to carry out further research work to address the following:

1. Comparisons between numerically-predicted weather inputs and actual measurements would establish confidence levels in the predicted values.
2. Development of a risk-based methodology to assess accuracy of numerically predicted or sensor measurement data before they are used in rating calculations
3. Application of ML algorithms and other post processing techniques to streamline the rating forecasting method developed in this project

Lessons Learnt

Key lessons learnt from the project are:

1. It is feasible to integrate high specification numerical weather forecasts into NGET’s rating model to generate an ensemble of OHL rating forecasts. Performance of this demonstrator system was validated against historical weather measurement data, ensuring expected behaviours and accurate outputs from the system.
2. Through validation of enhanced forecasts against historical measurements, several biases in weather forecasts were identified. It was found that these biases could be reduced using sophisticated post processing techniques on rating ensemble forecasts generated.
3. The change in approach from a single deterministic rating forecast to an ensemble forecast will enable NGET to choose the most suitable rating forecast for uplifts depending on weather conditions and network needs. A risk based assessment methodology will have to be established to make appropriate selection of rating forecast from the ensemble.
4. A detailed observation campaign would have to be carried out on sections of the network to enable a validation exercise and to gather inputs for a framework for appropriate selection of such enhanced ratings before embedding them into operational practice.

Dissemination
This project has been disseminated in the following way:

1. Conference: Talk in VIRTUAL RMetS Student and Early Career Scientists Conference 2021, 28th-29th June. (https://www.rmets.org/studentconf2021)
2. Inclusion in NGET Innovation Annual Summary 2020-21 (in preparation)

Following avenues are being considered for further dissemination:

1. Collaborative Project Summary and virtual webinar 
2. Publication in conferences/journals in meteorological and electrical domains