Niro designed and built five near-identical milk evaporators for Fonterra’s production site at Clandeboye and five at the Edendale site. Tubes in the Clandeboye evaporators often fouled excessively and occasionally blocked, requiring water blasting to clear them. Large amounts of undesirable foam from milk were observed in the second effect of some evaporators. This was known to be related to fouling and early shutdowns. The fouling increased the cleaning chemical and utility usage, and the evaporator downtime. The problems were believed to be related to the liquid distribution system at the top of the tubes. Evaporator liquid distribution has received relatively little research but it has been shown that the efficiency of evaporation reduces when there is poor wetting. Some estimates were available from previous work for the minimum flowrates required to obtain a complete falling film inside a tube. Many tasks were performed to assess the performance of the liquid distribution systems. The minimum wetting rates of three different milks were found in a model evaporator tube under isothermal, heat transfer and evaporation conditions at 60ºC. Numerous measurements of evaporators at Clandeboye were made to thoroughly analyse the evaporator performance. The overall heat transfer coefficients and wetting rates were calculated throughout the evaporators. Physical measurements were taken of the dimensions of the distribution systems in every evaporator. Many potential problems were found including warping, inconsistent hole sizes and fabrication faults. An analysis of the tube and distribution hole arrangements showed that every pass had some liquid misdistribution which was confirmed by a water trial. The evaporators were inspected before cleaning after 22 hours of whole milk production and after 5 hours of milk protein concentrate (MPC) production. There was considerable fouling at the bottom of some tubes that received low whole milk flows and large particles of MPC were blocking distribution plate holes. Observation and analysis showed that the foaming was likely to be caused by an upward flow of vapour from some passes which disrupted the downward flow of milk. As a result of this project there is sufficient confidence to justify modification of effects 2 and 4 of the evaporators. Sixteen tubes in effect 4 will be welded shut and the distribution systems in effects 2 and 4 will be redesigned to give a better liquid distribution. For effect 2, vertical tubes called ‘vapour risers’ should be installed to allow the vapour to flow upwards through the distribution plate without creating foam. The heights of the partitions dividing the effect 2 passes will be modified to encourage any foam in pass 1 to flow preferentially to pass 2. Installing a filter after the MPC direct steam injector will reduce the number of blocked distribution plate holes. Both modifications are expected to enable the evaporators to run continuously for 20 hours instead of 15 hours, giving up to 33% fewer cleans. The cleaning costs are approximately $700 and milk losses are approximately $200 per clean. The modifications should save up to $438,000, based on cleaning and water blasting in the 2003-2004 milk powder season. The design of future evaporator distribution systems must be improved to avoid retrofitting.