Rethinking fuel control
Author: Joe Peck
In this exclusive article for DCNN, Jeff Hamilton, Fuel Oil Team Manager at Preferred Utilities Manufacturing Corporation, explores how distributed control systems can enhance reliability, security, and scalability in critical backup fuel infrastructure:
Distributed architecture for resilient infrastructure
Uninterrupted power is non-negotiable for data centres to provide continuity through every possible scenario, from extreme weather events to grid instability in an ageing infrastructure. Generators, of course, are central to this resilience, but we must also consider the fuel storage infrastructure that powers them. The way the fuel is monitored, delivered, and secured by a control system ultimately determines whether a backup system succeeds or fails when it is needed most.
The risks of centralised control
A traditional fuel control system typically uses a centralised controller such as a programmable logic controller (PLC) to manage all components. The PLC coordinates data from sensors, controls pumps, logs events, and communicates with building automation systems. Often, this controller connects through hardwired, point-to-point circuits that span large distances throughout the facility.
This setup creates a couple of potential vulnerabilities:
1. If the central controller fails, the entire fuel system can be compromised. A wiring fault or software error may take down the full network of equipment it supports.
2. Cybersecurity is also a concern when using a centralised controller, especially if it’s connected to broader network infrastructure. A single breach can expose your entire system.
Whilst these vulnerabilities may be acceptable in some industrial situations, modern data centres demand more robust and secure solutions. Decentralisation in control architecture addresses these concerns.
Distributed logic and redundant communications
Next-generation fuel control systems are adopting architectures with distributed logic, meaning that control is no longer centralised in one location. Instead, each field controller—or “node”—has its own processor and local interface. These nodes operate autonomously, running dedicated programs for their assigned devices (such as tank level sensors or transfer pumps). These nodes then communicate with one another over redundant communication networks.
This peer-to-peer model eliminates the need for a master controller. If one node fails or if communication is interrupted, others continue operating without disruption. This means that pump operations, alarms, and safety protocols all remain active because each node has its own logic and control.
This model increases both uptime and safety; it also simplifies installation. Since each node handles its own logic and display, it needs far less wiring than centralised systems. Adding new equipment involves simply installing a new node and connecting it to the network, rather than overhauling the entire system.
Built-in cybersecurity through architecture
A system’s underlying architecture plays a key role in determining its vulnerability to cybersecurity hacks. Centralised systems can provide a single entry point to an entire system.
Distributed control architectures offer a fundamentally different security profile. Without a single controller, there is no single target. Each node operates independently and the communication network does not require internet-facing protocols. In some applications, distributed systems have even been configured to work in physical isolation, particularly where EMP protection is required.
Attackers seeking to disrupt operations would need to compromise multiple nodes simultaneously, a task substantially more difficult than targeting a central controller. Even if one segment is compromised or disabled, the rest of the system continues to function as designed. This creates a hardened, resilient infrastructure that aligns with zero-trust security principles.
Safety and redundancy by default
Of course, any fuel control system must not just be secure; it must also be safe. Distributed systems offer advantages here as well. Each node can be programmed with local safety interlocks. For example, if a tank level sensor detects overfill, the node managing that tank can shut off the pump without needing permission from a central controller.
Other safety features often include dual-pump rotation to prevent uneven wear, leak detection, and temperature or pressure monitoring with response actions. These processes run locally and independently. Even if communication between nodes is lost, the safety routines continue.
Additionally, touchscreens or displays on individual nodes allow on-site personnel to access diagnostics and system data from any node on the network. This visibility simplifies troubleshooting and provides more oversight of real-time conditions.
Scaling with confidence
Data centres require flexibility to grow and adapt. However, traditional control systems make changes like upgrading infrastructure, increasing power, and installing additional backup systems costly and complex, often requiring complete rewiring or reprogramming.
Distributed control systems make scaling more manageable. Adding a new generator or day tank, for example, involves connecting a new controller node and loading its program. Since each node contains its own logic and communicates over a shared network, the rest of the system continues operating during the upgrade. This minimises downtime and reduces installation costs.
Some systems even allow live diagnostics during commissioning, which can be particularly valuable when downtime is not an option.
A better approach for critical infrastructure
Data centres face incredible pressure to deliver continuous performance, efficiency, and resilience. Backup fuel systems are a vital part of this reliability strategy, but the way these systems are controlled and monitored is changing. Distributed control architectures offer a smarter, safer path forwards.
Preferred Utilities Manufacturing Corporation is committed to supporting data centres to better manage their critical operations. This commitment is reflected in products and solutions like its Preferred Fuel System Controller (FSC), a distributed control architecture that offers all the features described throughout this article, including redundant, masterless/node-based communication, providing secure, safe, and flexible fuel system control. With Preferred’s expertise, a distributed control architecture can be applied to system sizes ranging from 60 to 120 day tanks.
