Experimental measurement of cooling tower emissions using image processing of sensitive papers J. Ruiz a, * , A.S. Kaiser b , M. Ballesta a , A. Gil c , M. Lucas a a Departamento Ingeniería Mecánica y Energía, Universidad Miguel Hernández, Avda. de la Universidad, s/n, 03202 Elche, Spain b Departamento de Ingeniería Térmica y de Fluidos, Universidad Politécnica de Cartagena (Campus Muralla del Mar), Dr. Fleming, s/n, 30202 Cartagena, Spain c Departamento Ingeniería de Sistemas y Automática, Universidad Miguel Hernández, Avda. de la Universidad, s/n, 03202 Elche, Spain highlights < An improved digital image process in order to measure the emissions has been developed. < The RosineRammler distribution function do not provide consistent curve fits. < The Log-normal has proven to be the most suitable function for fitting experimental data among the ones tested. < AP-42 method overestimates the PM 10 emitted by the tower. < Drift, Total Dissolved Solids and the distribution of diameters (size and number) is required for a precise PM 10 evaluation. article info Article history: Received 19 July 2012 Received in revised form 30 November 2012 Accepted 5 December 2012 Keywords: Cooling tower emissions Sensitive paper Canny edge detector Log-normal distribution function abstract Cooling tower emissions are harmful for several reasons such as air polluting, wetting, icing and solid particle deposition, but mainly due to human health hazards (i.e. Legionella). There are several methods for measuring drift drops. This paper is focussed on the sensitive paper technique, which is suitable in low drift scenarios and real conditions. The lack of an automatic classification method motivated the development of a digital image process algorithm for the Sensitive Paper method. This paper presents a detailed description of this method, in which, drop-like elements are identified by means of the Canny edge detector combined with some morphological operations. Afterwards, the application of a J48 decision tree is proposed as one of the most relevant contributions. This classification method allows us to discern between stains whose origin is a drop and stains whose origin is not a drop. The method is applied to a real case and results are presented in terms of drift and PM 10 emissions. This involves the calculation of the main features of the droplet distribution at the cooling tower exit surface in terms of drop size distribution data, cumulative mass distribution curve and characteristic drop diameters. The Log-normal and the RosineRammler distribution functions have been fitted to the experimental data collected in the tests and it can been concluded that the first one is the most suitable for experimental data among the functions tested (whereas the second one is less suitable). Realistic PM 10 calculations include the measurement of drift emissions and Total Dissolved Solids as well as the size and number of drops. Results are compared to the method proposed by the U.S. Environmental Protection Agency assessing its overestimation. Drift emissions have found to be 0.0517% of the recirculating water, which is over the Spanish standards limit (0.05%). Ó 2012 Elsevier Ltd. All rights reserved. 1. Introduction Cooling systems have become essential in daily life. In fact, air conditioning is directly responsible for the increase in energy demand of the service industry. The increase of installed power for cooling system applications has lead to an increase in the consumption peak. Depending on the application, different cooling technologies should be applied in order to evacuate the heat of a refrigeration cycle. Among all the existing solutions, two types can be distinguished: those which employ atmospheric air as the condensation element (air condensation systems) and those which use recirculation water to accomplish the same task (evaporative cooling systems). The main difference between water and air condensation systems is that the condensing temperature and * Corresponding author. E-mail address: j.ruiz@umh.es (J. Ruiz). Contents lists available at SciVerse ScienceDirect Atmospheric Environment journal homepage: www.elsevier.com/locate/atmosenv 1352-2310/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.atmosenv.2012.12.014 Atmospheric Environment 69 (2013) 170e181