Improving Hydropower Safety and Efficiency
Statkraft Peru operates 18 hydropower plants in some of the country’s most rugged terrain, where water levels fluctuate based on rainfall and weather conditions. These facilities require constant monitoring for water availability and performance conditions, but traditional manual methods were labor-intensive and exposed staff to risky travel in remote locations.
To operate efficiently, plants like Cahua (299 GWh/year) and Malpaso (240 GWh/year) depend on accurate, timely data. Maintaining optimal reservoir levels is essential for efficient energy production, market timing, and grid stability.
Before modernization, each site had siloed systems and required operators to physically visit remote water sources to check conditions. Each hydropower station had its own isolated solution—SCADA systems, PLCs, or even Azure-based apps—making it hard to get a consolidated view into water resources for operations and market planning.
Operators sometimes had to travel five hours each way to collect a single water level reading. “We had to send a person just to get a value of the water level,” said Marcos Rodrigo De la Cruz Poblet, Data Engineer at Statkraft. “Now with HiveMQ we can get that value every minute without in-person inspection.”
Building a Centralized Data Backbone
Previously, stations sent data via FTP using a proprietary MIS file format. This approach was insecure and inflexible. “We were looking for a secure protocol,” said De la Cruz Poblet. “We chose MQTT over FTP because it supports TLS encryption and digital certificates, and HiveMQ supported all the security features we needed.”
Statkraft launched a new data acquisition system to consolidate real-time data from across all hydrometric stations. Many of these stations are located in mountainous areas with limited connectivity, making traditional communication protocols unreliable. The team selected MQTT as the protocol due to its lightweight payloads, efficiency in unreliable networks, and strong IoT support.
Statkraft chose HiveMQ Cloud because it offered the combination of MQTT, enterprise-grade reliability, and cloud-based flexibility. “The SLA was 99.95%, and HiveMQ was the only vendor we found that could meet our requirements,” said De la Cruz Poblet. “I also had experience with HiveMQ’s public broker from previous projects and knew it would work well in production.”
Statkraft subscribed to the HiveMQ Cloud Starter plan, which fits the use case because it supports dozens of remote stations without needing to send high-frequency data. “The water level on a lake doesn’t change every second. One data point every 30 minutes is enough,” he explained.
Real-Time Insights from Remote Lakes and Dams
Now, each hydrometric station sends MQTT messages to HiveMQ Cloud. A central FogLAMP server subscribes to the MQTT topics and transforms the messages into JSON before injecting them into both a local SCADA system and the global Aveva PI system, which is accessible via Grafana dashboards.
The collected signals—water depth, flow rate, sensor status, and battery voltage—are critical for downstream teams that depend on accurate, real-time insights. The Markets department is responsible for commercially optimizing and developing Statkraft’s portfolio to secure long-term value through global hedging strategies. Markets uses the data collected to model how much water to release and when, based on rainfall forecasts and energy pricing. The Operations team then executes on that plan.
“Markets goes to Operations and says, ‘Run the Chavez plant at 50% this week and 100% next week when prices are higher,’” said De la Cruz Poblet.
Improved Safety and Efficiency in Extreme Terrain
Today, the system contributes to safe, flexible operations across 450 MW of installed capacity and 2,500 GWh of renewable annual output in Peru. Many facilities are automated or semi-automated and located at elevations over 3,000 meters, remotely managed from a centralized dispatch center in Lima.
Beyond visibility and data consolidation, one of the biggest wins has been improving operator safety. “Some lakes are many hours away on dangerous roads. Now that we get real-time data, operators don’t have to travel every day,” said De la Cruz Poblet. “Statkraft is very committed to the safety of our people.”
Reducing operator travel to remote lakes means freeing up time and reducing risk. Real-time water level readings—now available every minute instead of once per day—support better forecasting and planning. These improvements are estimated to save Statkraft Peru more than $187,000 annually in reduced manual inspections alone, not including fuel, vehicle wear, or accident prevention.
The HiveMQ-powered system supports optimized water use across key hydropower plants like Cahua, Malpaso, and La Oroya, which collectively supply electricity to more than 400,000 homes.
Statkraft has integrated nearly all hydrometric stations into the system, with a few more planned. The lightweight MQTT protocol and HiveMQ Cloud plan provide ample capacity for current needs. “This project has solved a real business problem,” said De la Cruz Poblet. “It has improved operations, planning, and safety—all while giving us reliable, secure data we can trust.”
