Modelling wind speeds for cup anemometers mounted on opposite sides of a lattice tower: A case study Robert N. Farrugia a,n , Tonio Sant b a Department of Mechanical Engineering and Institute for Sustainable Energy, Room 217, Engineering Building, University of Malta, Msida MSD 2080, Malta b Department of Mechanical Engineering, Faculty of Engineering, University of Malta, Msida MSD 2080, Malta article info Article history: Received 18 April 2012 Received in revised form 9 November 2012 Accepted 28 November 2012 Available online 13 February 2013 Keywords: Cup anemometers Lattice tower Wind speed ratios Levenberg–Marquardt algorithm abstract When mounted at an intermediate monitoring level on a meteorological mast or tower, a cup anemometer’s horizontal wind speed reading is influenced by the structure’s construction type, the anemometer’s relative position to the structure itself and by the wind vector angle. Although measurements at the topmost level of the structure are a basic prerequisite for wind resource characterisation, intermediate height measurements are also required to calculate wind shear characteristics at the site of interest. Cup anemometers mounted at intermediate levels will, to a certain degree, always be influenced by the meteorological mast. Locating the sensors on opposite sides of the structure allows the wind analyst to filter the data according to wind direction, thereby reducing the largest interference effects. Failure of one of these same-level, opposite cup anemometers would curtail the possibility of data filtering by direction. A technique that would redress the consequences of cup anemometer breakdown during a measurement campaign was thus devised and validated. The capability of the Levenberg–Marquardt (LM) method to model data from one cup anemometer to the other sensor on the opposite side of the tower was assessed by comparing the modelled (LM-generated) data with measured data. The LM algorithm was then used to build a relationship between wind speed records for different concurrent and ever-increasing measurement programme time frames from the two active, same-level anemometers and then failure of one of the sensors was simulated. This paper presents results from a case study using 12 months of wind speed and direction data captured by two, same-level cup anemometers attached to opposite sides of a lattice-type telecommunications tower on the Mediterranean island of Malta. & 2013 Published by Elsevier Ltd. 1. Introduction The use of boom-mounted anemometers attached to tubular or lattice-type towers is a standard practice used by the wind industry for wind resource assessment. Round tubular masts and lattice towers are the most commonly-used wind measurement platforms; although both types interfere with the wind flow in their immediate vicinity exerting an influence on anemometer readings (Orlando, 2011; Kline, 2002; Lubitz, 2009). The installa- tion of new, dedicated wind monitoring structures at prospective wind farm sites could involve substantial capital costs and extend project time frames; especially in sensitive areas calling for permitting procedures. Therefore, the benefits afforded by already-existent high structures on prospective wind turbine installation sites should be considered prior to embarking on new wind resource assessment programmes. If structures such as telecommunications towers, cranes (Gouveia, Leach et al. 2007) and tower cranes, and power transmission pylons of suitable height could be used as wind measurement platforms, then this could reduce the capital outlay and waiting times associated with the procurement and permitting procedures required to install new, dedicated wind monitoring masts at the site of interest. Various studies have been undertaken to establish the level of influence of wind monitoring structures on the accuracy of wind speed measurements (Dabberdt, 1968; Hansen and Pedersen, 1999; Lubitz, 2009). Such studies aimed primarily to develop new models that predict the velocity deficit measured by a boom- mounted anemometer located a horizontal distance away from the centre of the tower. One study (Lubitz, 2009) proposed a model that used a potential flow solution in the region outside the tower wake zone and a Gaussian turbulent wake within the wake region. The effect of the tower structure on the distortion of the flow field was also modelled using computational fluid dynamics (CFD) techniques. A combined Navier–Stokes model and actuator disc model was applied by Hansen and Pedersen (1999) to investigate the impacts of point of measurement distance from the structure and mast solidity on the wind speed readings. IEC 61400-12-1 (IEC, 2005) presented plots of speed-ups and deficits around different mast and tower sections. The following Contents lists available at SciVerse ScienceDirect journal homepage: www.elsevier.com/locate/jweia Journal of Wind Engineering and Industrial Aerodynamics 0167-6105/$ - see front matter & 2013 Published by Elsevier Ltd. http://dx.doi.org/10.1016/j.jweia.2012.11.006 n Corresponding author. Tel.: þ356 23402577, þ356 21650675, þ356 21652249. E-mail addresses: robert.n.farrugia@um.edu.mt (R.N. Farrugia), tonio.sant@um.edu.mt (T. Sant). J. Wind Eng. Ind. Aerodyn. 115 (2013) 173–183