Increases in herd size and enforcement of water quality regulations have created an effluent disposal problem for the New Zealand dairy industry. Spray disposal to land and lagooning are commonly used but mechanical failures, management requirements and pressure on land have limited their suitability in many situations. This project was established to consider an alternative system. Initial studies revealed that anaerobic treatment in unmixed, non-insulated tanks, followed by trickling filter aeration, might be suitable. Two laboratory scale and one field treatment plant (1/15 - 1/20 full scale) were constructed to investigate the system. A factorial experimental design allowed investigation into three anaerobic treatment levels with a 3 x 3 aerobic treatment interaction nested within each anaerobic treatment. Anaerobic residence times of 5, 7.5 and 10 days provided loading rates of 1.35 - 0.63 kg COD/m3-day and 1.36 - 0.67 kg T S/m3-day. Removals between inlet and outlet averaged 71% and were insensitive to loading rate. Total solids accumulation rates of 40-50% TS input rate suggests that anaerobic tank design should be based on solids accumulation rate and cleaning frequency. The stone media trickling filter was loaded at approximately 0.61 kg COD/m3-day. Aeration periods of 1, 2 and 3 days and hydraulic loads of 2.8, 10.1 and 18.2 m3/m2-day were studied to determine their influence on treatment efficiency. Multiple regression analysis indicated that the longer residence times and higher recycle rates improved treatment efficiency. Removals varied with the measured parameters but ranged from 42-66% for COD. Design alterations to allow the final discharge to be taken from the bottom of the filter, after settling, would increase aerobic treatment efficiency above 75% COD removal. Prediction of treatment efficiencies beyond the monitored operating conditions suggested that only marginal improvements could be made. The TS accumulation rate in the aerobic phase was approximately 13% of the TS input rate or 56% of the BOD removal rate. Overall plant treatment efficiencies of 80-89% were obtained. Removals in excess of 92% could be achieved with minor design alterations. Maintenance and operational requirements were minimal. The only problem with the system was an average 15 fold increase in NO3-N and 4 fold increase in DIP under conditions for optimum removal of the other parameters. Intermittent land disposal could reduce this problem. Treatment comparison between similar laboratory plants, and between laboratory and field plants which varied by a scale factor of 56, suggests that identically designed plants would give a similar performance and that there is little scale effect. Increasing the scale only improved treatment efficiencies under unstable aerobic conditions, i.e., high recycle rates and low residence times. Increasing scale gave some decrease in maintenance and operational problems. Design of a full scale plant, based on daily pollution loads from a 250 cow dairy shed, suggests that the system is a viable proposition.

Possible areas of improving the efficiencies of the Lincoln University flame weeder are identified and investigated. The Hoffmann burner initially used in the Lincoln University flame weeder was found not to entrain sufficient air to allow complete combustion of the LPG used. A new burner, the Modified Lincoln University burner, was designed to improve the entrainment of air. Results show that the new design entrained sufficient air to theoretically allow complete combustion of the LPG, and this resulted in a 22.7% increase in heat output per Kg of LPG used over the Hoffmann burner. Temperature x time exposure constants required to kill weeds 0 - 15, 15 - 30, and 30 - 45 mm in size, were found to be respectively 750, 882, and 989 degrees Celsius.Seconds. These constants can be used to calculate the maximum speed of travel an operator can use a flame weeder at, once the temperature profile underneath its shields are established at various travel speeds, and therefore ensure that the flame weeder is used at its maximum efficiency. The constants can also be used to establish the cost efficiency of any flame weeder (in $/Ha), depending on the size of the weeds to be treated. The materials and methods used in establishing the temperature x time exposure constants can be used to establish the temperature x time exposure constant of any weed species at any size.

The present work is a contribution to description and understanding of the distribution and movement of water in swelling soils. In order to investigate the moisture distribution in swelling soils a detailed knowledge of volume change properties, flow characteristics and total potential of water in the soil is essential. Therefore, a possible volume change mechanism is first described by dividing the swelling soils into four categories and volume change of a swelling soil is measured under different overburden pressures. The measured and calculated (from volume change data) overburden potential components are used to check the validity of the derivation of a load factor, ∝. Moisture diffusivity in swelling soil under different overburden pressures is measured using Gardner's (1956) outflow method. Behaviour of equilibrium moisture profiles in swelling soils is theoretically explained, solving the differential equation by considering the physical variation of individual soil properties with moisture content and overburden pressure. Using the measured volume change data and moisture potentials under various overburden pressures, the behaviour of possible moisture profiles are described at equilibrium and under steady vertical flows in swelling soils. It is shown that high overburden pressures lead to soil water behaviour quite different from any previously reported.