International Conference Nuclear Energy for New Europe 2003 Portorož, Slovenia, September 8-11, 2003 http://www.drustvo-js.si/port2003 Preliminary Results of a Study on Hydrodynamics of Danube-Black Sea Channel Using Tritiated Wastewater from NPP Cernavoda Carmen Varlam, Ioan Stefanescu, Roxana Lazar, Mihai Varlam, Mihai Anghel National Institute R&D for Cryogenic and Isotopic Technologies- ICSI Rm. Valcea, P.O. Box 10, 1000 Rm. Valcea, Romania [email protected], [email protected], [email protected], [email protected], [email protected] Vasile Patrascu, “Grigore Antipa”National Institute for Marine Research, Mamaia Street, n03, 8700 Constanta, Romania [email protected] Cristina Bucur, Elena Bobric Environmental Laboratory of NPP- Cernavoda, Medgidiei Street, n01, 8625 Cernavoda, Romania [email protected], [email protected] ABSTRACT In this study we plan to use tritiated liquid effluents from a CANDU type nuclear power plant as a tracer, to study hydrodynamics on Danube-Black Sea Channel, thus extending the observation area along this important man-made channel. This channel is ideal for this kind of study, because wastewater evacuations are occasionally due to technical operations of a nuclear power plant. Tritiated water can be used to simulate the transport and dispersion of solutes in Danube-Black Sea Channel because they have the same physical characteristics as water. Measured tracer-response curves produced from injection of a known quantity of soluble tracer provide an efficient method of obtaining necessary data. This paper presents the mixing length calculation for particular conditions (lateral branch of the channel, and lateral injection of wastewater from nuclear power plant). A study of experimental published formulas was used to determine proper mixing length and the initial location of the experiment. Another result used in the further experiment concerns the tritium level along the Danube-Black Sea Channel. We measured tritium activity concentration in water sampled along the Channel between July-October 2002. The paper also presents future attempts to obtain unit-peak-attenuation (UPA) curve as related to different mixing times. 1 INTRODUCTION The hydrosphere is an important pathway through which pollutants can be dispersed into the environment. Regarding water quality control in rivers, both continuous pollution and accidental spills are the greatest ecological and economic potential dangers for the river. Numerous experimental studies in hydrosphere have been carried out to study the hydrodynamic behaviour of a wide type of stream. In most of these studies, chemical or fluorescent tracers are used, although they have various drawbacks, as they are usually nonperfectly conservative, their degradation products can be toxic and they are relatively expensive. The use of tritium as a tracer in river waters has several advantages [1]: 601.1 601.2 - it does not adsorb to sediments, being an ideal tracers as it forms water (HTO) molecules; - sampling is easy and it does not require special techniques; - small sample amounts can be used and extremely low tritium concentrations in water can be measured by low-background liquid scintillation; - the relatively long life of tritium permits storage before measurement. The location of Nuclear Power Plant Cernavoda near Danube-Black Sea Canal required developing a powerful program of environmental monitoring. Considering national and international regulations, interest area was formed around this important electrical energy provider. Using tritium routinely released as low-activity liquid radioactive waste by this CANDU type of nuclear power plant as a radiotracer, led to an extension of monitoring activity along the Danube - Black Sea Channel. We expect useful information about the flow characteristics in this important man-made channel at the end of this study. Preliminary results, about the baseline levels in interest area and the method used to establish the mixing length in this particular flow, are presented in this paper. 2 BACKGROUND AND TECHNIQUES Danube-Black Sea Canal is ideal for tracers’ study, because wastewater evacuations are occasionally due to technical operations of nuclear power plant. The possibility of a contamination agent being accidentally or intentionally spilled upstream from a water supply is a constant concern to those diverting and using water from this channel. A method of rapidly estimating traveltime or dispersion is needed for pollution control or warning the system on the channel where data are limited. Traveltime and mixing of water within a stream are basic streamflow characteristics needed in order to predict the rate of movement and dilution of pollutants that may be introduced in the stream. Tritiated water can be used to simulate the transport and dispersion of solutes in Danube-Black Sea Canal because they have the same physical characteristics as water, so understanding how tracers mix and disperse in a stream is essential to understanding their application in simulating pollution. 2.1 Theory of Transport and Dispersion for Soluble Tracers The dispersion and mixing of a tracer in a receiving stream takes place in all three dimensions of a channel. Vertical mixing is normally completed first, and lateral mixing later, depending on stream characteristics and velocity variations. Longitudinal dispersion, having no boundaries, continues indefinitely and is the dispersion component of primary interest. Thus, at section I, fig.1, vertical mixing could be complete, meaning that at any stream line and time, the tracer concentrations are the same throughout the water column, even though they vary drastically laterally. At short distance, section II, lateral mixing is still taking place, and the tracer mass in transport along the different stream lines is not equal because the response curves do not have equal areas. Mixing and dispersion in two dimensions, therefore, exist between sections I and III. An optimum mixing distance, section III, is reached when tracer response curves a, b, c,…, g, as observed laterally, have about the same areas, even though the individual response curves can vary considerably in shape and dimensions; dispersion is approaching the one- dimensional state. Nevertheless, the peak concentrations in the centre of the channel could be considerably greater than the peak concentration along the banks, while the latter response curves are longer both physically and in time of passage. Also, the tracer cloud is skewed, Proceedings of the International Conference Nuclear Energy for New Europe, Portorož, Slovenia, Sept. 8-11, 2003 601.3 Fig. 1 – Lateral mixing and longitudinal dispersion patterns and changes in distribution of concentration downstream from a centred injection of a tracer. [2] advancing faster in the centre of the flow compared with the channel boundaries. If the tracer curves are examined next at a long distance, the curve areas will be found to be nearly identical and concentrations nearly the same laterally. Thus, a dispersion state that is nearly one-dimensional exists between section III and IV, where longitudinal dispersion dominates. Ultimately, at extended distances, peak concentrations become virtually the same laterally, and longitudinal dispersion affects the shape and dimensions of the response curves exclusively. With time and distance, peak concentrations become attenuated and cloud lengths get longer and longer. One of the important assumptions of the tracer dilution principle in the third section is full mixing. In this case, and also for a continuous release under stationary hydraulic conditions, the tracer activity in the river (C0) can be calculated as: C C = t ⋅ q (1) 0 Q where Ct is the mean concentration of the tracer release, q the flow rate of the tracer discharge and Q the river discharge. This is valid for sampling points located at a distance from the source greater than the mixing length. A number of empirical formulae have been proposed to estimate the mixing length (L): - Day [3], proposed in 1977: L = 25w (2) where w is the channel width. - Guizerix and Florkovski [1], proposed in 1983: w 3 L = 10 (3) d where w and d are the stream mean width and depth, respectively. - Killpatrick and Cobb [2], proposed in 1985: vw 2 L = K (4) E z Proceedings of the International Conference Nuclear Energy for New Europe, Portorož, Slovenia, Sept. 8-11, 2003 601.4 where Ez is the transverse mixing coefficient, v is the mean channel velocity, w is the average channel width, and K is a coefficient depending on the degree of mixing, the location of the injection and the number of injections. A conventional manner of illustrating the response in a stream to a tracer is to plot concentration variation with elapsed time as observed at one or more points laterally in a stream. The shape and magnitude of observed tracer-response curves, figure 1, are determined by four factors: the quantity of tracer injected, the degree to which the tracer is conservative, the magnitude of the stream discharge, and longitudinal dispersion. Observed concentrations can be adjusted for the amount of tracer injected, for tracer loss and for channel discharge by use of so called “unit concentration”. Variations in dispersion on the same flow or different flows become most apparent if the unit concentrations for the peaks, Cup, are plotted as a function of elapsed time to the peaks. A plot of the peak concentrations (converted to unit-peak concentrations) with elapsed time for the response curves measured at four distances is