©2024 – Seeq Corporation | ‹#› ©2024 – Seeq Corporation | ‹#›

Presented by William Herewini Meridian Energy at Seeq Summit 2026 Scaling Condition-Based Monitoring through Data Democratization with Seeq MARCH 10th 2026

2

Presentation Agenda MARCH 10th 2026 | 3 Who are we?​

Our business challenge

​Our data pipeline

Use cases​

Results and benefits​ West Wind, Wellington

Harapaki Wind Farm, Hawke’s Bay

Who are we? MARCH 10th 2026 | 4 Aotearoa/New Zealand's largest energy generator​ with over 3000 MW of installed capacity equating to approx. 30% of the country’s electricity.

100% renewable generation – Wind, Water, and Sun​ 7 hydro stations 6 wind farms Grid-scale Battery Energy Storage Systems (BESS)

We retail electricity to more than 400,000 customers (or about 15% of household and business) across Aotearoa through our Meridian and Powershop brands.

Who are we? MARCH 10th 2026 | 5 The power to make a difference through data.

Improve processes from a routine-based maintenance approach to a data informed condition-based maintenance (CBM) approach.

Why? Make it easier for our on-site teams to do their jobs effectively. Improve asset health. Do our bit to keep the lights on in Kiwi homes and Aotearoa/New Zealand powered through the cold winter months.

Aviemore Hydro Station, Waitaki Valley Godley River delta, Lake Tekapo

Godley River Delta, Lake Tekapo

Our Business Challenge MARCH 10th 2026 | 6 Problem Technology gap in exploratory data analysis Insufficient data-driven information to inform important business decisions Complex data pipeline for CBM model development

Goal Meridian Energy would like to close the gap between our data team and our site crew, leveraging SME knowledge to accelerate the growth of our digital capability.

Move from routine-based maintenance to CBM, reducing operations and maintenance cost.

Aviemore Hydro Station, Waitaki Valley

6

Our Data Pipeline MARCH 10th 2026 | 7

AVEVA PI System collects data points from our assets and stores them within our PI Data Archive AVEVA PI Asset Framework and PI AF SDK builds analytical models to format and contextualise data AVEVA PI Vision constructs platform to collate and present relevant information from various data sources

White Hill Wind Farm, Southland

Manapōuri Hydro Station, Southland

Databricks builds models for forecasting and predicting events

Our Data Pipeline MARCH 10th 2026 | 8

White Hill Wind Farm, Southland

Manapōuri Hydro Station, Southland

Seeq offers a solution that caters to different levels of expertise, centralising the data, analytics, and visualisation in one place Seeq Organizer works well alongside existing tools to deliver updates, and report on significant changes

Use Cases MARCH 10th 2026 | 9 SF6 Monitoring and Alerting

Brake Service Forecasting

Turbine Bearing Oil Level Monitoring

Ōhau C Hydro Station, Mackenzie Basin

Use Cases MARCH 10th 2026 | 10 SF6 Monitoring and Alerting

Brake Service Forecasting

Turbine Bearing Oil Level Monitoring

Ōhau C Hydro Station, Mackenzie Basin

SF6 Monitoring and Alerting MARCH 10th 2026 | 11 Problem Though SF6 is an extremely efficient insulator for our circuit breakers, it is a potent greenhouse gas that must be monitored closely, as a leak would pose environmental risk.

Site crew perform biweekly monitoring of the pressure and temperature in the SF6 containers, reporting the readings to the engineering team to verify. Manapōuri Hydro Station, Southland

Manapōuri Hydro Station, Southland

White Hill Wind Farm, Southland

MARCH 10th 2026 | 12 Approach Install IIoT devices on one of our circuit breakers to capture pressure and temperature.

Utilise Seeq to replicate the work that is currently undertaken by our site crew.

Monitor SF6 pressure in Seeq, alerting on low pressure state, and high deviation between phases.

Ingest alerts via Databricks, forwarding the data to PowerBI and PowerApps for user interaction. Manapōuri Hydro Station, Southland

Manapōuri Hydro Station, Southland

SF6 Monitoring and Alerting

White Hill Wind Farm, Southland

MARCH 10th 2026 | 13 Fig: SF6 model in Seeq, alerting during maintenance period. SF6 Model in Seeq

TESTING PERIOD

MARCH 10th 2026 | 14 Alerting Channel POC

Use Cases MARCH 10th 2026 | 15 SF6 Monitoring and Alerting

Brake Service Forecasting

Turbine Bearing Oil Level Monitoring

Ōhau C Hydro Station, Mackenzie Basin

Brake Service Forecasting MARCH 10th 2026 | 16 Problem Brake servicing relies on routine-based maintenance.

There is difficulty in converting to maintenance decisions based off model output due to the disconnect between the model and SMEs.

We risk over or under-servicing our assets when we rely on routine-based maintenance.

Benmore Hydro Station, Waitaki Valley

MARCH 10th 2026 | 17 Approach Find an interim approach to data-driven maintenance by triggering maintenance based on brake operations.

Use Seeq to monitor the operations of our brakes, forecasting when we expect maintenance to be due.

Benmore Hydro Station, Waitaki Valley Brake Service Forecasting

MARCH 10th 2026 | 18 Fig: Brake operation forecasting model in Seeq.

Brake Service Forecasting Model in Seeq TEST THRESHOLD

Use Cases MARCH 10th 2026 | 19 SF6 Monitoring and Alerting

Brake Service Forecasting

Turbine Bearing Oil Level Monitoring

Ōhau C Hydro Station, Mackenzie Basin

Turbine Bearing Oil Level Monitoring MARCH 10th 2026 | 20 Problem Site crew must perform manual checks to ensure there are no oil leaks, and that the oil level is within expected bounds.

There can be difficulties with visual inspection while the unit is online, due to confined spaces and large moving parts.

Benmore Hydro Station, Waitaki Valley

MARCH 10th 2026 | 21 Approach Use Seeq and existing sensor data to calculate the daily rate of change.

Alert when this metric is outside of normal operational ranges, based on historic data.

Note: We are only interested in slow leaks, as anything sudden will be addressed by generation controllers.

Benmore Hydro Station, Waitaki Valley Turbine Bearing Oil Level Monitoring

MARCH 10th 2026 | 22 Fig: POC turbine oil level model in Seeq. Oil Leak Model in Seeq

Results and Benefits MARCH 10th 2026 | 23 Closing the Gap​ Seeq’s low-code/no-code environment is appealing to those who don’t have the time to learn complex tools, while providing functionality that other tools cannot provide.

Trust in Models The Seeq UI provides accessible views of the data and analytics that are involved when developing models.​

A Large Step towards CBM By centralizing the data, analytics, and visualisation, Seeq makes it easy for users to pick up and create CBM models. This aids us in moving current processes from a routine approach to a data-driven approach.

West Wind, Mākara

Meridian Energy has efficiently adopted Seeq, enabling data democratisation, and accelerating the growth in our digital environment MARCH 10th 2026 | 24 Challenge Technology gap in exploratory data analysis Insufficient data-driven information to inform important business decisions Complex data pipeline for CBM model development

Solution Rolling out and utilising Seeq to cater to different levels of expertise. Develop clear model solutions, mimicking existing processes such that SMEs can build trust in the models.

Results Low-code/No-code analytics environment with full functionality. Clear UI and views into how models operate. Development of models to replace routine tasks with CBM.

Ōhau A Hydro Station, Mackenzie Basin

Looking Ahead MARCH 10th 2026 | 25 Continue to on-board users around Meridian, increasing data capability and fostering a Seeq user community.

Incorporate Seeq into our daily workflow.

Replace routine maintenance on one hydro unit with CBM models.

Using Seeq’s Data Lab, develop more complex models to gain insights, and add-on develop tools for users to use and explore.

Waitaki Power Station, Lake Waitaki

Questions? MARCH 10th 2026 | 26

Te Āpiti wind farm, Manawatū

William Herewini​ Meridian Energy @: will.herewini@meridianenergy.co.nz Engineering Data Analyst

image1.jpeg

image2.jpeg

image3.png

image4.jpeg

image38.png

image39.png

image26.emf

image27.png

image28.png

image29.png

image40.jpeg

image41.png

image42.jpeg

image43.png

image44.png

image45.png

image46.png

image47.png

image48.png

image49.png

image50.png

image51.png

image52.jpeg

image53.png

image54.png

image55.jpeg

image56.png

image57.png

image58.jpeg

image59.jpeg

image60.png

image61.jpeg

image62.jpeg

image63.jpeg