Chaffey Dam Project - CRCFE projects B.202 and B.203 / MDFRC project M/03/5066
CRCFE Technical Report
Despite a significant research effort into understanding the physics of artificial reservoir destratification both overseas and in Australia, little work has been done regarding chemical and biological measures of the effectiveness of destratification. Historically, destratification systems have been designed to reduce the temperature difference between the top and bottom of the water column without specific consideration of their chemical or biological effects. Designers have implicitly assumed that a reduction in the temperature change would lead to complete mixing of the water column. This mixing commonly is assumed to produce a deeper surface mixed layer (SML) in the reservoir, and to increase dissolved oxygen concentrations in the deeper waters. Increased dissolved oxygen concentrations lead to decreased concentrations of soluble iron and manganese and decrease the release of nutrients from the sediment. However, close inspection of field data offered as proof of successful destratification (Le. approximately uniform temperature in the vertical) invariably reveals a remnant temperature stratification near the water surface.Artificial destratification may control blue-green algal blooms in two ways. By deepening the surface mixed layer, it reduces the amount of light available to an algal cell and hence its rate of photosynthesis. The dominant bloom-forming algal species in Chaffey Reservoir (discussed in Chapter 9), Anabaena and Ceratium, do not grow well under well-mixed, Le. lower light, conditions and so a deeper SML may exert a selective pressure that favours other species. By reducing the amount of dissolved nutrients, especially phosphorus, in the water column, destratification can reduce the amount of algae that can exist in the reservoir. As all types of algae require phosphorus to grow, a reduction in reservoir phosphorus content will reduce the total amount of all types of algae within the reservoir without necessarily changing the distribution of algal species. The Destratification and Water Quality project addressed several significant knowledge gaps regarding the chemical and biological responses of a reservoir, Chaffey Dam near Tamworth, NSW, as it underwent artificial destratification. Specifically, we sought to understand: which processes limit the supply of oxygen to the hypolimnion and sediments - how phytoplankton respond physiologically to sudden changes in the ambient light and nutrient environments • how sediment nutrient releases are linked to changes in the temperature and oxygen concentration at the sediment -water interface - how phosphorus is transformed between different size fractions and chemical species - the impact of dessication, arising from large changes in reservoir level, on sediment nutrient dynamics - the quantities of nutrient supplied to the reservoir from internal (sediment) and external (catchment) sources Routine field sampling took place from September 1995 to June 1997. This sampling program supplemented the existing monitoring performed by the NSW DLWC. Ten intensive field experiments of 7-10 days duration were undertaken to examine short -term effects of artificial destratification. In addition, DLWC provided its entire historical water quality database for the reservoir and catchment so that data collected by the project could be placed in a historical context. Significant findings - The 100 L s" destratification system at Chaffey Dam was operated on five occasions between February 1995 and April 1997. Each period was of 4 - 5 weeks duration except April 1997 when the system was turned off after 10 days because of extreme oxygen depletion near the surface of the reservoir. The destratifier did not cause complete mixing of the water column at any time, even when the reservoir was at 20% of capacity nor did it produce a deeper surface mixing layer (SML). Despite producing a weaker temperature gradient, there was insufficient mixing energy input at the water surface via surface heat losses and wind stirring to deepen the surface layer (Section 7.2). Microstructure temperature profiles showed that temperature changes as small as 0.05 °C were sufficient to prevent downwards transport of buoyant blue-green algae. After 14 days of compressor operation the vertical temperature distribution reached a quasi-equilibrium state after which continued operation provided little or no further reduction in the temperature change across the water column. A quasi-equilibrium oxygen distribution was established typically after 3 weeks of operation. The quasi-equilibrium condition took the form of nearly linear temperature and oxygen gradients and represented a balance between sources of heat and oxygen near the surface of the reservoir, downward transport due to large-scale destratification-induced circulation, and, in the case of oxygen, in situ consumption throughout the water column. Theoretical design considerations indicate that a 5-fold increase in airflow rate (and number of plumes) would be able to destratify a full reservoir with a 16°C temperature difference in about three weeks. However, numerical modelling results using meteorological data from the reservoir required an airflow rate of up to 1800 L s-1, to alter the SML sufficiently to reduce the growth of phytoplankton through light limitation. The difference between the theoretical (700 L s") and numerically-derived (1800 L s") design airflow rates reflects the importance of local climatological conditions which are considered only in the numerical simulations. When commenced in spring, destratification maintained oxygen concentrations at 50-60% of saturation at the bottom of the water column. The elevated oxygen concentrations removed iron, manganese and hydrogen sulfides within a few days. As soluble iron was oxidised, dissolved phosphorus was bound to iron complexes and became unavailable to the phytoplankton. Destratification reduced the net amount of phosphorus released from the sediments to the hypolimnion by up to 85%. The Peel River is a major source of nutrients to Chaffey Reservoir (Section 8.3.7). Delivery of nutrients from the catchment typically exceeded the observed net changes in reservoir nutrient content implying that the reservoir acts as a sink for the external nutrient load. In the Peel R, both turbidity and suspended sediment were poorly correlated with FRP and TP and reasonably well correlated with nitrate and total kjeldahl nitrogen. On average, 30% of the total phosphorus carried by the Peel River was present as filterable reactive phosphorus. The ratio of FRP:TP did not correlate significantly with gauge height. The dominant source of the sediments (and presumably the particulate phosphorus) carried by the Peel R is known to be the eastern catchment (Section 8.2.3). However, the source of FRP within the catchment is unknown because geochemical tracing techniques consider only particulate matter. FRP delivered by the Peel R accounted for at least half of the observed change in total phosphorus content in the reservoir during the study period. This result implies that during normal-wet precipitation years, the catchment may supply 50% or more of the phosphorus that is subsequently converted into algal biomass. The vast majority of dissolved phosphorus in the water column was present as orthophosphate, the most bioavailable form. Following an initial spring or early summer phytoplankton bloom which eventually sedimented out of the surface layer, concentrations of phosphorus were very low, approaching the limit of detection. Low phosphorus and nitrogen concentrations were often present but seldom reduced the cell division rate. However they did prevent the accumulation of algal biomass until autumnal deepening of the surface layer entrained hypolimnetic nutrients into the surface layer. The cell division rate determines how quickly the maximum biomass is attained. FRP accumulation in the hypolimnion was the major determinant of the following year's algal biomass but did not impact on species dominance. The phytoplankton community was dominated (88% of the algal biomass) by either motile or positively buoyant species that remained in the surface layer. The surface layer was rarely deeper than the euphotic depth and so provided an excellent light environment for algal growth. Blue-green algae have grown in all months of the year at Chaffey Dam.
Algal management implications - Destratification may prove to be effective in reducing the algal biomass present in the reservoir because it effectively reduces the internal phosphorus load from the reservoir's sediments. It should also reduce the amount of particle-bound phosphorus delivered by the Peel R that is recycled into a bioavailable form. However, the inability of destratification to deepen the SML, and thereby change the light environment, makes it ineffective for the elimination of blue-green algae at Chaffey Dam. Alternative strategies for maintaining oxic conditions in the hypolimnion should be considered. For example, direct injection of pure oxygen into the hypolimnion could be just as effective at reducing the internal nutrient load while at the same time preserving the thermal stratification. This has the advantage of keeping the nutrients lost from the surface layer through particle sedimentation in the hypolimnion and ultimately the sediments. This strategy should produce lower algal biomass in summer and autumn than would destratification which resuspends some of the nutrients in the water column. ncreased emphasis on tracing catchment FRP as opposed to TP sources is required for informed catchment management decision making. A large proportion of TP (30%), is delivered by the Peel River as FRP. Geochemical sediment tracing techniques cannot identify the source of FRP within the catchment. Unequivocal identification of the catchment FRP source(s) will require the deployment of streamflow guages and stage-dependent samplers on the tributaries to the Peel upstream of Taroona in order to compute nutrient loads. An earlier attempt by DLWC to identify these sources using 'rising stage' samplers did not provide data of sufficient quality to estimate loads. ere is insufficient Fe available to scavenge all the FRP in the water column when the destratifier is operating. Operation of the destratifier converts reduced iron (FeII) with a low affinity to bind phosphorus to iron oxyhydroxide (FellIOOH) which has a high phosphorus binding affinity. The imbalance between Fe and FRP increases the importance of controlling FRP delivery from the catchment. Manganese is often present in higher concentrations than Fe in the anoxic hypolimnion but it does not playa significant role in P reduction. Dessication of dam sediments decreased the capacity of sediments to both take up P, and to release P under anoxic conditions. The duration of this effect is unknown but is believed to depend on the time required for the anaerobic bacteria to recolonize the sediment following reinundation. It may be possible to reduce the internal P load by operating the dam to cause low water levels of sufficient duration to dessicate a significant portion of the sediments. Extensive sulphate reduction occurs during summer stratification and this contributes a significant portion of the internal phosphorus load. Maintenance of aerobic conditions within the hypolimnion may prevent sulphate reduction and the attendant release of phosphorus. Elbow-grab water samples gave higher biomass estimates than 5 m-integrated samples, especially during periods of blue-green algal dominance. We recommend the pooling of 5 m-integrated samples from 5 locations in the reservoir into a single sample for algal abundance estimates. This will provide a more representative estimate of reservoir conditions. However, if contact with algal scums is an issue, then surface samples need to be collected as well.
MDFRC funding agency: Cooperative Research Centre for Freshwater Ecology
MDFRC client: Cooperative Research Centre for Freshwater Ecology