Skip navigation
The Australian National University

Modelling

To understand how solar cells work it is necessary to model them theoretically. We do that using advanced 3D simulation programs, such as Sentaurus Device, as well as in-house tools such as QS Cell & Quokka.

QS Cell

QS Cell is a versatile program for the simulation of several characterisation techniques commonly used to determine the electronic properties of silicon wafers. In particular, it can model QSSPC measurements of the effective carrier lifetime, Suns-Voc curves, photoconductance-based measurements of recombination currents at diffused regions of silicon wafers, and photoluminescence spectra. The program can also model complete I-V characteristics of solar cells, their fill factor, efficiency and spectral response. It is applicable to arbitrary carrier injection conditions and to compensated doping. Collaborations in this area include Dr Keith McIntosh.

Download QS Cell 6.0.

Download QSCell PRC for partial-rear contact simulation

QSS Model.

Figure 1: A screenshot of the main sheet of the QS Cell program developed at ANU. In this example, a typical p-type cell at its maximum power point is simulated.

More details can be found in the following publications:

  • A. Cuevas and R. Sinton, "Simple modelling of solar cells", Proc. 23rd European Photovoltaic Solar Energy Conf., Valencia, September 2008, pp. 315-319.
  • Andres Cuevas, "Modelling silicon characterisation", Energy Procedia 8, 94-99 (2011).

Contacts

Andres Cuevas
Kean Chern Fong - kean.fong@anu.edu.au

Quokka

In 2012 the 1D/2D/3D steady state solar cell simulator Quokka was developed by Andreas Fell [1]. It aims for fast simulation speed and high user friendliness, and is available together with a web-based input file generator on PVLighthouse.com.au. Quokka solves the general semiconductor carrier transport equations based on the quasi-Fermi potentials in a simplified way. Those simplifications are namely conductive boundaries and quasi-neutrality, which are well-known to be valid for most typical silicon solar cell devices.

The conductive boundary approximation excludes highly doped surface regions from the solution domain. This greatly relaxes the need for fine meshing to resolve the doping profile and the space charge regions. The doped surfaces are rather modelled by a surface recombination current, typically defined by a recombination current pre-factor J0, as well as by lateral current transport within the boundary with the characteristic sheet resistance Rsh. Despite the boost in simulation speed due to the coarser mesh, J0e and Rsh are practically the most relevant input parameters for solar cell modelling, as they can be straightforward derived from experimental characterization.

The quasi-neutrality approximation assumes the electron and hole concentrations to be equal within the solution domain, i.e. the bulk. Quokka utilizes this by solving the set of semiconductor equations reduced to the two charge carrier transport equations only. This greatly improves numerical stability and the overall simulation speed.

The derived steady state model is implemented by a finite volume approach in MatlabĀ®, utilizing a powerful inbuilt algorithm to solve the resulting nonlinear set of algebraic equations in a fast way. A license-free compiled version is available on PVLighthouse.com.au. Some main features of the released software are:

  • Pre-defined layout for a partial rear contact (PRC) and all back contact (ABC) unit cell, with a variety of settable contact /doping patterns and dimensions
  • Fully automated meshing with user-defined quality
  • Generation profile import from OPAL2 (advanced optics calculator on PVLighthouse.com.au)
  • Automated IV-curve sweep, open circuit and maximum power point finding
  • Full loss analysis (e.g. at maximum power point ) listing all recombination and resistive loss contributions (by free energy loss analysis, FELA)

Advantages of Quokka compared to the state-of-the-art device simulator Sentaurus is the simulation speed, free availability, user friendliness and more practically relevant input parameters for doped surfaces. Other available similar implementations of the conductive boundary approach are PC2D, which is much slower and inflexible regarding the geometry, and CoBoGUI, which requires a COMSOLĀ® license, both of them are currently restricted to two dimensions.

Contacts

Andreas Fell - andreas.fell@anu.edu.au

Updated:  4 February 2014 / Responsible Officer:  JavaScript must be enabled to display this email address. / Page Contact:  JavaScript must be enabled to display this email address. / Powered by: Snorkel 1.4