OZ-ENERGY-ANALYSIS.ORG   -   open science for the new millennium

THE STORIES   |   DATA   |   ANALYSIS   |   MODELS   |   LITERATURE   |   DISCUSSIONS

Status: [24th Oct 2010] Round One

DEMAND MANAGEMENT

Introduction and Scope

Demand Management can refer to different things, but mostly means reducing the peak demand. Sometimes called peak shaving or load shifting, electricity is used now instead of later, and/or later instead of now. Off-peak hot water (for those without solar hot water) is a simple example of this. By flattening (and fattening) demand peaks the required upper capacity of system components is held down.

While the wholesale electricity market reflects demand and supply dynamics by pricing electricity in half hour chunks (allowing the price to rise as high as $12,500 a MWh from an average of around $25 a MWh), there is no Time of Use (ToU) price signal flowing through to residential consumers.

Here we are especially concerned with "Direct Load Control", whereby loads and appliances are actively controlled over minutes and hours in order to better manage the electricity system. While issues of efficiency (and the longer term issue of building design) are part of 'managing' demand, they are not part of the scope here. Similarly the term 'Smart Grid' can encompass various aspects, but here we confine attention to metering and Direct Load Control.

This is introductory material, an evolving primer to help in making sense of more specific and technical documents produced by government, utilities, and others. The focus here is on the residential sector, which constitutes around one quarter of demand on average but a much larger portion of peak demand, due in large part to air conditioners. For now we take it that the implementation of demand management in the commercial sector will be much the same as for residential. The industrial sector is a separate issue and not considered here.

Teasing out possibility from policy, reckonings from rollout, is an ongoing process. We are especially interested in what IS happening, or being considered at a policy level, and the development of this page over time will reflect that focus.

The important question for OzEA is this: how much peak shaving and load shifting will the demand management systems of the future provide?

Renewables

Demand Management can help in running an electricity system whether or not renewables are involved, however, in the current context it can only help around the edges with the large scale (multi-day) variability in renewable power sources, especially Wind.

That said, it is none-the-less the case that a system incorporating more renewables needs to counterbalance with increased demand side flexibility. What we discuss here is the first generation of demand management, and how this might play out into the future remains unclear. Certainly, Direct Load Control can act to balance short term fluctuations in supply and thus enhance system stability.

Metering and the 'Smart' Revolution

'Smart Meters' are a key component of a 'Smart Grid', but what does a Smart Meter do?

Interval Meters are the first step. While a traditional meter simply measures overall electricity use, and is read every three months, an interval meter records the electricity used each half hour (or other time interval). They also include communication capabilities that make usage data available to the electricity provider, and also the consumer. This may be simply informative to start, but Time of Use billing is the usual goal. Residential consumers will then pay less at times when supply is plentiful, and electricity will be more expensive at peak times. Victoria is in the process (2009-13) of an Interval Meter rollout.

There seems to be ongoing debate as to precisely what functionalities should ideally be included in a smart meter rollout, and we avoid delving into these details just yet.

While time-of-use electricity pricing better apportions cost towards those who contribute most to peak loads, and can provide incentive to change usage behaviour, these consequences in themselves do not constitute 'Direct Load Management'. What is required is for specific appliances to be managed in an automated way in response to the relative abundance or scarcity of supply. There are different ways this can happen, but first we have a look at the target appliances.

Appliances

Only some appliances have flexible usage patterns; things like kettles and computers do not. One particular group of appliances is of particular interest, being the triumvirate of air conditioners, water heaters and pool pumps. Each of these is a fixed-in-place item that can have a significant role in demand management.

The 'off peak' hot water tariff was an early and basic form of Demand Management, and water heating (so far as it consumes electricity) continues to allow load shifting. Water heating appliances can mostly be switched off at times of peak demand and otherwise run intelligently to minimise the cost of the electricity consumed.

Pool pumps simply should not be on at times of peak load, and with consideration for noise issues can be intelligently scheduled to run on the cheapest electricity.

Air conditioners are not so trivially time shifted, but as the system-crippling load at peak times they need to be included in the base level of Demand Management systems. It is apparent that two levels of control are needed. First, a consumer subject to time-of-use pricing and empowered with automated control tools can choose to reduce air conditioner use at peak times through either turning them off (one by one in households with multiple air-conditioners), and/or by moving the thermostat up and back down in response to the price highs. Second, as an agreed part of ones electricity plan (or as an emergency measure), the air-con compressor (but not the fan) can be switched off for short periods according to some schedule; thus providing a mechanism for centralised load curtailment at times of extreme system load.

Fridges are another major appliance that could act as a managed load. A standard example is for a fridge to slightly lower its thermostat (e.g. 3.5 deg C instead of 4.0 deg C) when electricity is plentiful / cheap (or in advance of peak load), and to become just a little less cold (e.g. thermostat at 4.5 deg C) when demand is high. Fridges also have significant potential to utilise thermal mass; this can allow both load shifting in response to the electricity price, and for fridges to do their work cycles more efficiently in the cooler times (e.g. night and early morning).

Load control -is not- achieved by bluntly cutting the power in and out; rather, these devices are to have embedded circuitry that controls the device in response to communication signals. In the case of Air Conditioners there is an Australian Standard (AS4755.3.1) for the control interface. As these standards are established, the inclusion of the necessary 'chip' in new appliances can be mandated.

Implementation: How will it work?

One way it could work, and a good starting point for this discussion, is via a Control System that is independent of the meter. Such a Home Area Network (HAN) would manage and communicate with the appliances; it would also need to communicate externally to obtain information about the electricity price. The HAN might exist as an internet device on your home network, and by default communicate with, and be managed by, your electricity company. Your choice of electricity plan might specify how you want the appliances managed. Alternatively, you might want to manage your HAN in some other way, with your electricity company simply billing you.

A separate HAN means that rollout of a basic Interval Meter can proceed, and consumers can upgrade their HANs over time as appliances change and other technologies become important. Alternatively, rolling the HAN into a Smart Meter means working out what needs to be in the Meter and getting it right once-and-for-all before any rollout. The aim here would be an essentially seamless transition to a smarter grid, and one that avoided the hassle of having and managing a separate HAN.

The astute reader may be bursting with practical questions, as we are here at OzEA, but this is a conceptual discussion for now. Clarity will develop in time, while further immediate progress appears to require wading into an abyss of detail.

Residential Time of Use Pricing

Residential Time of Use (ToU) pricing will be the motivator that drives households and businesses to adopt usage behaviour that will provides demand side elasticity. It will fundamentally connect the supply side to the demand side. It will drive the industries that will provide the technologies for load management.

It is important to realise that households with modest electricity use at peak times may very well end up with smaller electricity bills, effectively withdrawing subsidy to those who contribute most to the peak loads. Running multiple cheap air conditioners will become a less and less attractive way of making up for poor house design.

A long lead in may be required to implement ToU pricing, for two reasons. First, the social and political aspects require community acceptance; second, ToU pricing without commensurate (and automated) load control systems will likely struggle to maintain community acceptance.

In September 2010 the Energy Networks Association (ENA) (industry association made up of Australia's energy distribution and transmission businesses) released a strategy for the implementation of smart electricity networks (4 page summary [1.7 MB], 20 page full report [5.2 MB]). While this strategy is somewhat coy in referring only to "price structures", it is clear that Time of Use pricing is a central aspect.

Concluding Remarks

Utilities, governments and commercial providers are active in developing the technologies and policies to implement demand management as an integral part of the electricity system. In some jurisdictions (e.g. Victoria) smart meter rollout is happening now.

Load management (and efficiency) are important for managing peak demands, especially from air conditioners, and are needed to avoid or postpone expensive upgrades to distribution networks in particular. Also, incorporating renewable sources of power into the generation mix (especially wind) adds a new layer of 'peakiness' into the task of having supply meet demand at all times. Demand-side management has an important albeit secondary role in managing the variability of renewables.

We maintain a watching brief on this area, and for now look to the comments below as the active arena. At this time we are more concerned with what Demand Management can do for us than vice versa. In ten years time, say, how much peak shaving and load shifting capability will these technologies provide compared to a 2009 baseline? At some point in 2011 we expect to need an estimate of this.

[Show Full Lists]

 

  TOP   |   REFRESH PAGE   |   DISCUSSIONS

fc - Oct 2010