Background
District Heating (DH) systems will invariably be contaminated by bacteria. 90% of all bacteria in such technical systems are attached to surfaces in contact with water. When immobilised on such surfaces, bacteria form biofilms. Inside biofilms the chemistry is significantly different than in the bulk of water. Formation of biofilms, also called biofouling, on metal surfaces may lead to corrosion of the metal.

The resulting corrosion, termed microbiologically influenced corrosion (MIC), is localised, develops fast and can cause severe reduction in system lifetime; e.g. through wall pitting of SS 316 at a rate of approx. 0.5 mm/year has been observed in a Danish peak load system. Similarly, the combined production of heat and electricity has necessitated the use of storage tanks, which have been shown to be especially vulnerable to MIC. Biofouling may also lead to increased drag in pipes thus increasing the costs for pumping, decreased heat transfer in heat exchangers, and clogging of filters and monitoring equipment.

A preliminary survey on biofouling, corrosion, and corrosion mechanisms has been performed throughout selected DH plants within Hungary , Czech Republic , Sweden , Finland , Denmark , Germany , Austria , and Great Britain . Performing this survey had two purposes.: 1) to find out if MIC currently experienced in Danish DH systems may be threatening the integrity of other DH plants and 2) to establish a connection between MIC and system parameters generally available in order to asses the risk of MIC in general.

In summary, general corrosion rates ranged from 1 to 40 micrometer/year, biofouling from 10 3 to 10 7 cells/cm 2 , and local corrosion rates from 0 to 400 micrometer/year.

The corrosion within two out of the ten plants were categorised as MIC. Four plants experienced corrosion that were categorised as either "possibly MIC" (clearly other factors influencing the corrosion though all prerequisites for MIC were present) or "initiating MIC" (weak indications for MIC). In the remaining four plants either no corrosion was observed or the corrosion was clearly not influenced by microbiological activity.

Comparing the observations of MIC and non-MIC with plant parameters obtained during the survey, gave some indications of what could lead to MIC and what should not. Particularly the use of chemical additives seemed to play a major role in MIC. The following was observed as indications:

• Addition of sulphite as oxygen scavenger increases the risk of MIC
• Addition of hydrazine lowers the risk of MIC but other corrosion problems may occur.
• Keeping a relative low temperature results in higher risk of MIC
• Not maintaining a sufficiently high pH increases the risk of MIC
• Addition of phosphate buffers the pH and lowers the risk of MIC
• Addition of ammonium has similar effects and lowers the risk of MIC

Conclusions
MIC is not only a Danish problem, but potentially a problem to all DH installations. We recommend that micro organisms and the problems they obviously create in many plants are taken seriously.

On identification and monitoring of MIC we recommend that corrosion coupons are used either directly within the system or within a side stream exposure unit like the pipe flow unit used in this study. For corrosion measurements we recommend that measurements of general corrosion rates are accompanied by topographical analysis of local corrosion since the local rates are often much higher than the general ones. Growth based methods for monitoring biofouling have not proven to be useful. It is instead recommended to use total counts for determining the level of biofouling. The by-product of the corrosive SRB bacteria, sulphide, may be identified through a simple spot test for sulphide rich minerals. It is also recommended to use x-ray fluorescence in the case detailed investigations of possible MIC attacks are needed.

Prepared by
Danish Technological Institute, Mr Bo Højris Olesen

Danish Technological Institute, Environmental and Water Technology

Swedish Corrosion Institute