Research article A multivariate statistical analysis to evaluate and predict ignition quality of marine diesel fuel distillates from their physical properties D. Karonis ⁎, I. Zahos-Siagos, D. Filimon, F. Vasileiou National Technical University of Athens, Greece abstract article info Article history: Received 22 February 2017 Received in revised form 22 May 2017 Accepted 19 June 2017 Available online xxxx Ignition quality of diesel fuels used in compression ignition engines is mainly described by their respective Cetane Number (CN). The most widely accepted method for CN measurement is the ASTM D613, involving a variable compression ratio Cooperative Fuel Research (CFR) engine. However in the recent years other alternative methods based on a Constant Volume Combustion Chamber (CVCC) concept have gained significant popularity mainly due to lower cost of acquisition, maintenance and operation, still being able to provide accurate results. Other widely accepted parameters for the characterization of ignition quality of diesel fuels are cetane indices. These are empirical mathematical equations, based on simple physicochemical properties (such as density and distillation characteristics) and usually correlate well with CN. However, they are accurate enough within narrow limits and are very prone to produce misleading results when involving modern fuels, or biofuels and their blends. In this work, an effort was made to investigate the room for improvement for the prediction of CN, apart from the use of the most widely accepted method of Calculated Cetane Index (CCI), standardized by ASTM D4737, with regard to today's used marine diesel fuels. A total of 47 distillate marine diesel fuels (with high sulfur content compared to automotive diesel fuels) has been employed in order to create multivariate re- gression models to predict CN equivalents. Thus, useful conclusions were drown in an attempt to evaluate the im- portance of various physicochemical properties with regard to ignition quality, as well as a first step to the potentiality of revaluation of empirical mathematical models currently in use. © 2017 Published by Elsevier B.V. Keywords: Cetane number Marine diesel fuel Aromatics content Multiple linear regression 1. Introduction Cetane Number (CN), one of the most substantial properties of a die- sel fuel, is used to indicate its behavior right after the injection inside the cylinder of a compression ignition engine, therefore its ignition quality. It is usually estimated by the time delay period between the start of in- jection and the start of combustion, generally referred to as Ignition Delay (ID) [1,2]. ID depends on many factors related to the engine itself, such as com- pression ratio, injection pressure, injection nozzle, cylinder and piston geometry, but also the fuel. To be able to deliver high ignition quality, a diesel fuel usually has to perform short ID times. The shorter the ID, the less premixed fuel and air accumulate inside the cylinder before combustion [1–6]. Thus, when combustion finally takes place, the rate of burning remains within reasonable limits, avoiding too high pressure rise rates that increase noise, decrease efficiency and can damage the engine [1,3–7]. It is clear, then, that a high CN indicates short ID, for a given fuel [1]. Furthermore, fuels with low ignition quality characteris- tics exhibit poor exhaust emissions performance [1,2,8–10]. To measure the ignition quality of a given fuel typically a Coopera- tive Fuel Research (CFR) engine is used, as mandated by ASTM D613 method [11]. This procedure is however time consuming, requires skilled personnel and large amounts of expensive chemicals as stan- dards as well as reference and sample fuels, let alone the high expense of acquiring and maintaining a CFR engine [12]. These factors have led to the development of alternative methods to estimate CN, some of them involving instrumentation built around a Constant Volume Combustion Chamber (CVCC) where the sample fuel is actually burned [6,13], and some of them employing empirical equa- tions that use commonly measured physicochemical properties of the fuel to determine the ignition quality of the fuel [6,12,14–20]. The idea behind CVCC methods is to simulate the prevailing condi- tions inside a compression ignition engine, right before the fuel injec- tion, then inject the sample fuel and measure the ID observed during the auto-ignition process. CVCC methods are not a new concept. The idea to use a CVCC to measure the ID of a diesel fuel goes back to 1930′s [6]. These methods have evolved through years of research and development to become nowadays the most popular experimental pro- cedures for the evaluation of a fuel's ignition quality, due to more Fuel Processing Technology 166 (2017) 299–311 ⁎ Corresponding author at: School of Chemical Engineering, National Technical University of Athens, Iroon Polytechneiou 9, Zografou, 15780 Athens, Laboratory Z.202, Office A.302, Greece. E-mail address: dkaronis@central.ntua.gr (D. Karonis). http://dx.doi.org/10.1016/j.fuproc.2017.06.021 0378-3820/© 2017 Published by Elsevier B.V. Contents lists available at ScienceDirect Fuel Processing Technology journal homepage: www.elsevier.com/locate/fuproc