Bord na Mona has a problem with a manufacturing by-product that affects all its product lines - water. The peat has just too much of it and it has to be expelled before the peat becomes useable.

The sun starts the process during harvesting but this is not enough. Getting rid of the remaining water costs money, both as energy needed for heat drying or as extra labour. Keeping these costs to a minimum is therefore important for the company.

A student research project has resulted in an automated moisture measuring device that will help Bord na Mona do just that. It provides an accurate indication of percentage moisture content and can help reduce unnecessary handling.

The level of water left in the peat is critical when manufacturing briquettes, said Dr William O'Connor, a lecturer in the Department of Mechanical Engineering at University College, Dublin. Too much and the peat sticks in the molds, too little and it is too dry to compress into a briquette.

A colleague, Dr Donal Finn, is involved in a collaborative applied research project with the peat company. It is investigating methods of milled peat drying and has developed a pilot system.

Dr O'Connor is an expert in the development of sensor devices. When he heard that an automated sensor device was needed to optimise the drying process, Dr O'Connor decided it would make a good student project and assigned it to fourth year, Mr John Holland, to it.

"Peat comes off the bog mostly water. To use it for most purposes you have to dry it," Dr O'Connor said. "Part of the difficulty is the stuff coming in is so variable." Its density, water content and handling characteristics change, even within a relatively small harvest area.

Peat for briquettes is dried in a five stage process, he said. After milling it is passed through elaborate thermal dryers to drive off moisture which drops from about 65 per cent down to a narrow target around 10 per cent.

"There can be a fair bit of wastage in the beginning. If they get it wrong they don't briquette it. There is a fair bit of trial and error to get it right."

Existing "gravimetric" methods involve weighing, drying and weighing again to calculate the correct moisture level. The method is slow and labour intensive and gives only a poor indication of what the overall batch is like. "If they had an inline moisture sensor they would be in a much better position to automate it and to optimise the system," Dr O'Connor said.

He started with a "blank sheet" and looked at a whole range of options including improved gravimetric, ultrasound, conductivity and resistance. "Then we focused on microwaves."

Mr Holland developed a system that uses microwaves to give a continuous measure of moisture content. "It was an outstanding project," Dr O'Connor said.

Measuring moisture in a variable product was a challenge, he said. "We need two measurements, moisture and total mass." The microwaves give both in a continuous stream of information.

Water in the peat takes some of the energy out of the microwave signal and the system measures this microwave "attenuation". Density or mass can be interpreted by looking for changes in other aspects of the wave, Dr O'Connor explained. The system provides a continuous stream of information, giving a much more accurate picture of the peat batch and telling the operators when the peat is optimal for briquette production.

"There are many industries that need a reliable moisture sensor," he said. Brewers and distillers measure moisture in grain, coal is monitored for water content and water is important in woodchip production and in certain foodstuffs.

Linking the sensor to a computer system would open up many possibilities, Dr O'Connor said. "It is quite easy to make it do other things" such as turn up or down heating systems, monitor and control quality, channel product into storage or production and so forth.

The measuring system was good enough to win Mr Holland the Hewlett-Packard Award 2000, a competition open to all third-level institutions "for an outstanding project in engineering, science or information technology".

Another student, Mr Roger Kissane, is now refining the system, simplifying the microwave source and measuring equipment to make it cheaper. The researchers were surprised that the technology was not already well established in industry. "That is what we asked ourselves. Why hasn't this been done before if it works."

Looking back through the literature they found that a Polish researcher had developed a similar system some years ago to measure water in coal and woodchip but this wasn't refined or brought to market. Later they found some commercial devices incorporating some aspects of the ideas.

Dr O'Connor has no immediate plans to commercialise the system, preferring instead to continue his research work on sensor development. He has designed systems for sensing fluid density, measuring the fluid level in beer kegs, sensing very low accelerations and a device that can count people. The UCD group is willing, however, to license novel aspects of the technology so that some other group could commercialise the device. It does seem like an opportunity for an instrument company looking for new markets.