Poul Harremoës

Genetic algorithms in real time control applied to minimize transient pollution from urban wastewater systems

Real time control aims at optimization of the urban wastewater system performance under dynamic loading from rain. This paper presents a novel approach to control the whole system: sewer system, treatment plant and receiving water with the aim to achieve minimum effects of pollution. The application of nonlinear model predictive control by means of a genetic algorithm reveals excellent results with hypothetical problem sets. The methodology makes it possible to optimize the system performance directly with respect to water quality parameters and to avoid the traditional indirect and artificial performance criteria, such as permissible annual overflow volume. The relevance of this novel approach is illustrated by the fact that no stringent correlation has been found in the investigation between the reduction of overflow volume and the increase of oxygen concentration in the receiving water.

pH-decrease in nitrifying biofilms

Abstract Nitrification is an acidity producing process. It has been shown theoretically that the diffusional resistance to the transport in the biofilm of the inorganic carbon species as affected by the acidity production in a nitrifying biofilm gives rise to a decreased pH in the interior of the biofilm. These theoretical results have been verified on biofilms developed on the surfaces of a rotating drum under well controlled laboratory conditions. The results show clearly the drop in pH as predicted by theory. The phenomenon can give rise to unexpected effects on the performance of nitrifying biofilms, when most of the bacteria work under a much lower pH than the pH measured in the bulk water.

Stochastic models for estimation of extreme pollution from urban runoff

Abstract The calculation and evaluation of urban runoff has to be shaped according to the pollution problem in question. It is important to group such problems into two categories: acute and accumulative. The first calls for extreme statistics on an event basis, the latter on an annual basis. Deterministic modelling of pollutant transport during rain fails to eliminate very significant residual variability and stochastic elements have to be introduced. It is shown that accounting for the variability corresponds to an increase in pollutant transport for extreme events by a factor of 2–3. This ought to be of significance to the design of combined sewer overflows.

Immediate and delayed oxygen depletion in rivers

Abstract It is important to distinguish between those processes that remove organic matter from the water phase of a river with an immediate oxygen consumption, as opposed to those processes that remove the organic matter by fixation at the bottom where it will be degraded with delayed consumption. An investigation of a steady-state discharge shows the importance of this distinction, but a theoretical analysis shows this to be of far greater importance during and after an abrupt discharge. It is particularly true in case of storm overflows from combined sewers—as demonstrated experimentally in this and the following paper.

Nitrification in rotating disc systems—I: Criteria for transition from oxygen to ammonia rate limitation

Abstract Identification of the rate limiting substrate and half-order rate constants are presented on the basis of biofilm kinetics for triple substrate conditions (organic matter, ammonia and oxygen) in a rotating biological disc system. These findings have been verified by pilot-scale experiments. The transition from ammonia rate limitation to oxygen rate limitation occurs at a bulk ammonia-nitrogen concentration to bulk oxygen concentration ratio of approx. 0.4. Accordingly, oxygen is the rate limiting substrate. For most practical purposes the experiments showed that the oxygen limited half-order reaction constant in the presence of simultaneous organic matter degradation is reduced compared to nitrification alone by a factor which is a function of the growth of the heterotrophs and the nitrifiers in the biofilm.

Risk terminology—a platform for common understanding and better communication

The sciences analyzing and describing risks are relatively new and developing, and the associated terminologies are developing as well. This has led to ambiguity in the use of terms, both between different risk sciences and between the different parties involved in risk debates. Only recently, major vocabularies have been compiled by authoritative agencies. Some of these vocabularies are examined and explained based on a division into fundamental and action oriented risk terms. Fundamental terms are associated with description and characterization of the chemical, biological and physical processes leading from risk source(s) to possible consequences/effects. The approach to these terms is based on a cause-effect skeleton. The action oriented terms cover administrative, scientific, sociological, etc. processes associated with the work of identifying, characterizing, regulating and communicating risks in the society, and their internal connection and iterative character have been illustrated. Focus is laid on engineering and toxicological risks, but to some extent, the thoughts presented may be extrapolated to other areas. Differences in applied terminology probably cannot be eliminated, but they can be identified and clarified for better understanding. With the present paper, the authors hope to contribute to reducing the probability of derailing risk discussions from the risk issue itself.

Degradation mechanisms of colloidal organic matter in biofilm reactors

Abstract The degradation mechanisms of colloidal organic matter in biofilm reactors have been studied in an idealized laboratory reactor system with soluble starch as a model substrate. Batch tests and experiments with different reactor configurations have shown that for this specific substrate, bulk liquid hydrolysis is the mechanism for transforming non-diffusible organic matter into biofilm diffusible substrate. A simplified mathematical description has led to the identification of the degree of hydrolysis, D H , as the parameter expressing the major difference between degradation of diffusible and non-diffusible organic matter in a biofilm reactor. D H depends on the combined volumetric and surface hydraulic loading rate, Q 2 /(AV). In full-scale wastewater treatment plants, the degradation mechanism presented in this paper can explain important differences between the performance of trickling filters and RBC reactors.

Oxygen and pH fluctuations in rivers

Abstract A mathematical model for oxygen fluctuations in a river reach, involving photosynthesis, total community respiration, and reaeration is presented. Analogy is established between diurnal oxygen fluctuations and diurnal pH-fluctuations. Comparison is made between oxygen production and respiration, and carbon dioxide consumption and production. Twin curve analysis of diurnal oxygen fluctuations and diurnal pH-fluctuations is made of measurements on a single river reach in Denmark. The measurements were carried out about once a month for a little more than one year. The analysis has revealed seasonal variations of the river parameters: total daily gross production of oxygen, total community respiration and reaeration coefficient. These parameters are important for river modelling.

Estimation of regional intensity-duration-frequency curves for extreme precipitation

Regional estimation of extreme precipitation from a high resolution rain gauge network in Denmark is considered. The applied extreme value model is based on the partial duration series (PDS) approach in which all events above a certain threshold level are modelled. For a preliminary assessment of regional homogeneity and identification of a proper regional distribution L-moment analysis is applied. To analyse the regional variability in more detail, a generalised least squares regression analysis is carried out that relates the PDS model parameters to climatic and physiographic characteristics. The analysis reveals that the mean annual number of extreme events varies significantly within the region, and a large part of this variability can be explained by the mean annual rainfall. The mean value of the exceedance magnitudes can be assumed constant for intensities with durations less than one hour. For larger durations a pronounced metropolitan effect is evident, the mean intensities in the Copenhagen area being significantly larger than found in the rest of the country. With respect to second and higher order moments the region can be considered homogeneous for intensities with durations less than 24 hours. A regional parent distribution is identified as the generalised Pareto distribution.

On the potential of genetic algorithms in urban drainage modeling

Abstract A genetic algorithm (GA) is a stochastic search algorithm that applies the biological concept of survival of the fittest in order to search for the optimal solution to a problem. In this paper we explore the potential and the benefit of using GAs for solving problems in urban drainage modeling. The main problem areas where such methods are assumed to have some benefit as compared to traditional procedures are identified from the literature as model calibration and model predictive control. The use of GAs for multi-criteria decision analysis is not reported in the context of urban drainage modeling but believed to be an interesting field of application. The methodology is discussed by means of benchmark problem sets for each of the applications.