Fuel cell control systems

by Mark Wilkinson

The principle of the fuel cell is known to most schoolchildren, but despite its discovery 165 years ago, the technology is still in its infancy for automotive applications. One key challenge is the control needed to make it suitable for powering road vehicles.

The first fuel cells produced just a few watts, whereas the automotive industry needs tens of kilowatts to motivate its vehicles. Today’s automotive fuel cells require a whole ‘army’ of peripheral equipment to make sure that they run reliably, safely and powerfully. As a static system, the fuel cell is complicated enough to control; but when one tries to apply this technology to the automobile, all the parameters of running that highly dynamic platform complicate the issue even further. A powerful electronic control system becomes essential.
Just like the internal combustion engine, the fuel cell needs its own ECU; taking in multiple data inputs to respond immediately to driver demands whilst maximizing efficiency. As Marc Wiseman, Ricardo manager in the US, puts it: “Control strategies for fuel cells are a ‘hot area’ – everyone is trying to maximize the performance available from these devices. Core to this process is understanding how to control the fuel cells.”

One might imagine that with only heat and water as by-products of the electricity generation process, the fuel cell ECU could be simpler than that needed for an IC engine, but it is actually even more of a challenge. As Professor Robert Thring, Intelligent Energy Chair of Fuel Cell Engineering (and head of Aeronautical and Automotive Engineering at Loughborough University, UK) explains; “The requirement for electronic control in fuel cell vehicles is at least as demanding as in conventional piston-engined vehicles.”

Efficiency, but with rapid response, too

The traction PEM fuel cell vehicle is currently limited to just a few hundred miles’ range due to the limited amount of hydrogen that can be safely stored on board, so efficiency is extremely important. However, rapid response is just as important; the road vehicle is a highly dynamic device; the driver won’t be prepared to wait whilst the extra power is generated. In car applications, the power has to be available immediately, particularly for situations like the cut-and-thrust of rush-hour traffic.

Start-up times are a also sensitive issue for the fuel cell vehicle; Charlie Wartnaby, chief engineer of leading UK electronic control specialist, Pi Technology, explains that Ford’s 2002 fuel cell Focus uses a standard ignition key, but when starting, the vehicle warns the driver that only limited power will be available for a short period of time. He says that starting is complex to orchestrate and that it takes up to two minutes for the stack to be brought up to full output.

Fuel cell control is complex

Efficiency and responsiveness aren’t the only concerns, though; the control system needs to be fully aware, too. Dennis Hayter, Business Development Manager for Intelligent Energy, explains; “The system needs to control water and heat management as well as all its own internal requirements and external power demands. It also needs to be very aware of its surroundings, constantly questioning them with internal logic algorithms, such as: “Are there other cells around me?” “Are they all balanced?” “Do I sense hydrogen - if not, should I initiate an emergency shut-down?” “What do I feed the electric current into – and how should I do that?” “Am I running at the right temperature or do I need to be cooled down?””

For instance, when the driver requires more power, the compressor has to force more air through the stack to increase the oxygen supply. However, cooling does not always need to be constant – at times, it may be better to turn down the cooling effect so that it consumes the least possible energy. According to Qinetiq, public technology development arm of the UK’s Ministry of Defence, there are complex trade-offs between fuel economy and performance. For example, to respond instantly to a steep increase in power demand, the stack needs high pressure air and an excess of hydrogen within each cell, but if this was maintained at all times the fuel economy would be very poor. On the other hand, if high fuel economy were the goal, the stack would be slow to respond and would probably require a battery to deliver short bursts of power.

Ron Hodkinson, managing director of Fuel Cell Control adds: “Controlling a fuel cell is like conducting an orchestra, it depends on the system, but you need a controlled supply of gases and fluids through the fuel cell stack, whatever the technology used. You need this not only in the operating condition (once the cell has begun operation), but there are also all the issues of getting it going in the first place and then closing it down afterwards.”

Humidity control is critical

If the PEM membrane dries out, efficiency tails off dramatically - effective humidity management is critical to the life of the stack. Hodkinson continues:

“With PEMs, water management is a serious issue all of the time… the hydrogen needs to be wet, and you need to control the humidity to between, typically, 75 and 80% for it to work properly.“

Hodkinson explains that there are two main types of PEM solutions for traction applications; load-following, which uses the stack to produce all the energy all of the time, from idling in town to full-bore acceleration; and the battery-assist, which uses the stack for average power demand - the battery or accumulator helps out during times of peak power demand (hill climbing, rapid acceleration, etc). “Humidity is one of the key technical challenges, because when you use a load-following technique there is a control issue in trying to keep the humidity correct when the load is changing very rapidly. What tends to happen is that you either flood or starve the system, and this is one of the things that causes short life in fuel cells.”

Safety is a major issue

The control system needs sensors to report on the hydrogen supply at all times - not just for economy, but also for safety. One insider notes that today’s fuel cell control system also include a wide range of protection strategies; H2 sensing, leak detection, venting and purging functions. There is a feeling of caution where hydrogen safety is concerned; an hydrogen ‘event’ could be extremely damaging to the public image, and therefore, future, of the fuel cell-powered vehicle. In fact, so many protection measures are taken that some could easily become redundant in the future.

Confidentiality is paramount

The OEMs are all working under extreme secrecy, unwilling to outsource the development of any core technology. Developers are invited to work on various aspects of the fuel cell system, but the OEM brings these separate aspects together and keeps the vital control algorithms to itself. For instance, Pi Technology developed the software for three ECUs in Ford’s 2002 fuel cell Focus (Vehicle Systems controller, Thermal Systems Controller and the Energy Management Module). The secrecy surrounding the project was evident – that vehicle had a dozen major ECUs, all communicating on a LAN, but Pi didn’t ever get to see the final integration, which was kept strictly in-house by Ford.

Outside the big OEMs, however, the budgets are lower, and partnerships have formed to manage the development of FC projects for smaller-volume applications such as buses. The Fuel Cell Propulsion Institute’s program to develop fuel cell-powered mining vehicles is a typical example. In this multi-player project, Ricardo has taken the lead role; in addition to developing the control system, it oversees the integration of the system components into the final vehicle.

Generic electronics but unique algorithms

At the heart of each FCCS is a programmable logic controller ‘black box’ from electronics suppliers such as Bosch, Mitsubishi or Siemens. Each developer uses one or more of these black boxes, but programmed with its own unique fuel cell control algorithms. Ballard, for example, uses Bosch controllers, but it is Ballard’s own algorithms that make it unique.

Primary controller requirements:
• Safety – checking for hydrogen leaks
• Durability – shutting down if contaminants are detected, “contaminants can kill the stack really fast,” says Quantum Technologies. A 99.99%-pure, or better, supply of H2 is essential
• Efficiency – to maximize range of limited on-board hydrogen
• Cell health – every cell in the stack needs to be monitored: voltage, current, temperature and pressure of H2 and O2