1 Introduction

Whilst much is said about the effect of renewable energy intermittency on electricity networks, the information shared and views expressed are often anecdotal, difficult to verify and limited to a particular technical, geographical or social context. There is surprisingly very little real-world data on the effects of intermittency on electricity networks, particularly in regard to solar intermittency. Investigations into the issue in Australia are particularly scarce for both photovoltaics (PV) and concentrating solar thermal (CST), especially when considering unique network aspects such as the distance covered by Australia’s electrical infrastructure and its limited level of interconnection (see Figure 1). Analysis of the effects of solar intermittency and operational considerations are vital for accelerating solar technology commercialisation and deployment in Australia.

The largest concentration of research into the effect of intermittent renewable generation (IRG) on electricity networks has been in the wind industry, with one of the most widely cited works being that of the Institute of Electrical and Electronics Engineers (IEEE) Accommodating Wind’s Natural Behaviour - Advances in Insights and Methods for Wind Plant Integration [1]. The issues raised in such wind industry research papers include, but are not limited to: dispatchability, balancing, variability, uncertainty, market operation and impact on system reliability. There is no consideration of whether these issues apply to solar generation and if so, how they relate and the extent of the commonality.

Recognising this general lack of information, this report provides an in-depth analysis of worldwide research and practical results on renewable generation intermittency, examining what common conclusions can be made from other efforts in this area, and how these might apply in the Australian context. Some of the most significant work in this area is being undertaken by the California Solar Initiative Research, Development, Deployment and Demonstration (CSI RD&D) Programme, exploring the ‘planning and modelling for high penetration PV on the California transmission and distribution network’[2]. This work is specifically targeted towards California, and only examines photovoltaic generation. More broadly, the current worldwide state of the art in solar intermittency study is represented by the information being collated by the International Energy Agency (IEA) Task 14 on high penetration PV. This work is predominantly general in nature, collating already available data rather than conducting targeted experiments, and again focuses on PV intermittency. The US Department of Energy (DoE) is also facilitating many activities to investigate high solar penetration issues. Their High Penetration Solar Portal [3] contains links to information on various case studies, investigations, analysis and technical discussions on relevant high solar penetration topics.

Figure 1 Unique network aspects in Australia

*Australian load profiles and environmental conditions such as fire and flood

There is no common definition of high penetration intermittent generation, but there is consensus amongst the parties developing the IEA Task 14 that a high penetration situation exists if additional efforts are necessary to optimally integrate the dispersed generators. Building on this philosophy, a working definition of high penetration intermittent generation was developed and is presented in Section 2 of this report.

Section 3 summarises the current state of research on renewable generation intermittency, focussing on centralised and distributed photovoltaics (PV) and Concentrating Solar Thermal (CST) although wind intermittency is also addressed briefly. The literature shows that one of the main challenges for the power network is overcoming issues with the instantaneous penetration of intermittent solar generation, that is the fraction of total system load being provided by solar generation at a given instant. Solar generation is viewed as negative load and when it is combined with the actual system load yields a ‘net load’ which corresponds to the power that must be supplied by other resources on the system. The effect of renewable generation intermittency on net load variability, as covered in the existing literature, is also given in this section. Observations of actual cases of high penetration solar intermittency are also presented. Section 3.3 summarises studies examining the likely impacts of increased levels of intermittent renewable generation through simulation and modelling. Changes to operations, infrastructure, planning and power quality management that might be required to manage increased levels of intermittent renewable generation, as suggested in the literature, are discussed in Section 3.4.

To understand the differences between wind and solar intermittency, several aspects of both wind and solar generation intermittency are discussed in Section 4. This includes correlation of both generation types with load, their inherent variability over different timescales, and load-following and regulation requirements to meet the new variability in net load introduced by solar and wind. The differences between CST and PV power intermittency are discussed in Section 5, which includes comparisons of the two different solar generation types.

Section 6 presents studies completed in other countries which are considered relevant for informing the integration of high penetration intermittent generation in Australia. Both system and distribution level studies are included. Findings from some of these studies can be applied in Australia, but similar studies would need to be carried out in an Australian context, examining unique Australian conditions, such as reduced grid interconnection, disparate population densities and long and skinny feeders, before any significant conclusions can be made here.

This project builds upon the early work of the ASI-supported Australian PV Association’s International Energy Agency (IEA) Task 14 investigation into High Penetration PV in Electricity Grids, by investigating CST as well as PV. Section 7 of this report discusses key intermittency issues attributed to PV installations in various countries around the world, identified at the IEA Task 14 High Penetration Photovoltaics workshops in December 2010 and May 2011.

Section 8 of this report explores solar intermittency in the operation of Australian electricity networks. An Australian industry workshop and follow-up survey were conducted to obtain the views of key solar industry experts on solar intermittency in the Australian context. Invited attendees and submissions included responses from Australian utilities, power system operators, large-scale renewable system operators, renewable energy developers and other solar and electricity industry players, on solar intermittency in the Australian context. The discussions and findings from both the workshop and survey are presented in this Section.

Apart from the daily sun cycle, clouds are the main cause of solar generation intermittency. Variations in irradiance depend on cloud height, sun elevation and wind speed. Irradiance fluctuations have been observed and analysed in Germany [7], Japan [8], Belgium [9] and USA [66], but analyses of power output fluctuations are particularly scarce. Ten months of 10-second resolution solar data from the Desert Knowledge Australia Solar Centre (DKASC) in Alice Springs, Australia, was collected and analysed to evaluate the occurrences of power fluctuations and the various ramp rates investigated. To study the variations of output power from a small-scale and large-scale PV plant, 5-second data from the CSIRO Energy Centre’s office building rooftop PV system and 1-minute data from Australia’s largest flat-panel PV system at the University of Queensland were also collected and analysed. Section 9 of this report presents an analysis of the various PV plant output power ramp rates observed and considers the frequency of their occurrence. The data for line-to-neutral AC voltage measured at the DKASC were also analysed and the correlation between variations in irradiance, output power and voltage was observed.

The effects of intermittent power generation upon electricity networks are strongly influenced by the type of network to which they are connected and the type and amount of generation installed, however most fall into the categories of stability and voltage effects. A simulation model developed at CSIRO was used to examine the likely impacts of output power fluctuations seen at the DKASC on various types of Australian electricity networks with different penetration levels of solar power. Four different scenarios comprising weak and strong grids with high and low level penetration of solar power were modelled. Section 10 of this report explores the likely effects of output power fluctuations seen at the DKASC on various Australian electricity networks.

The power generated by solar power plants has an intermittent character due mainly to atmospheric effects such as insolation variability resulting from cloud movement, however differences are present between thermal and PV technologies and their response to changes in insolation. One example involves the generally low inertia to change in output of PV plant when compared to thermal systems, which have a level of inherent thermal storage, while the reliance of concentrating solar thermal (and PV) systems on direct irradiance can be contrasted to non-concentrating PV’s acceptance of both global and direct irradiance. Further, consideration of the effects of shading and partial shading indicates that the larger the PV plant, the longer it takes for a cloud cover spread to shade a significant proportion of the entire field.

An analytic model has been developed in which the PV plant power output is described as the signal output of a first order low-pass filter whose input signal is the solar irradiance signal. Using historical or predicted irradiance levels, the developed model, explained in Section 11, can be used to simulate and predict the power output of an existing or proposed PV plant. This allows estimation of the output power ramp rate probability density function, which can be used to estimate the effect of a particular PV array upon the local network.