21. Enlightened Energy Management

Enlightened Energy Management is not just about turning off unnecessary lights
-- it is about lightening the load on the electric grid using all your smartness.

We wake up to a tune played by our smart phones, immediately click on various apps just in case there’s an alert or a message that can’t  wait, use the tap (aka faucet) at full flow, drive to work in a car even when public transport is easily available, enter an office pre-cooled to a low temperature that we are forced to wear a jacket, and so continues  our day -- consuming more energy than we need, with consequences that we have little time to ponder.

But, why should an individual care? Consider a typical home electricity bill. For many, it is a small fraction of all the monthly expenses. The usual perception is that even when we diligently turn OFF unnecessary appliances at home, the electricity bill is unlikely to drop by more than the  cost of a meal in a mid-scale restaurant. But, given its multiplicative effect, such diligence can help  power deficient countries like India,  whose citizens are forced to contend with power of poor quality (occurrences of blackouts, brownouts and load shedding).

Avoiding or reducing energy usage, by turning OFF devices when not needed, is obviously a highly desirable goal. The multiplicative effect of simple automated steps  can go a long way in addressing the energy problem. But, rather than depending on humans to intervene, there is need  for automated means to manage the devices to reduce energy consumption and these must be exploited especially for  Heating Ventilation and Air Conditioning (HVAC) systems  since they are responsible for a large proportion (30-50%) of energy consumed by buildings.  They should be designed to use easily available/ sensed information to control HVACs and other devices, like fans and lights, without compromising on the thermal comfort. Automated approaches are especially needed to ensure energy conservation, when it comes to large scale and distributed operations so as to make a difference not only to the purse but also to the environment.

So, what is needed for this automation? A lot of research has been done on how energy consumption of a building is correlated to occupancy, number of appliances, temperature and other environmental factors. For example, various sensors such as motion, door or pressure sensors, cameras, etc. can be used to detect occupancy. Temperature of the building can be observed using temperature sensors and power consumption can be measured using smart meters and clamp-on meters.  These sensors enable automated central administration to reduce and optimize power consumption, while also remotely keeping a check on the health of the energy consuming  appliances, maintaining quality of the atmosphere, and tracking occupants during different times of the day (useful to drive staffing and air conditioning based on footfall), to name a few.

With connectivity, infrastructure, and hardware platforms becoming easier to manage than before, prototypes and proof of  concept deployments exist in research labs around the world, including ours at IIT Bombay, and scores of companies that have begun to empower electricity consumers with automated systems that help them better govern energy consumption and achieve significant reductions in energy bills. These use  Internet of Things  (IoT)  based products to do inexpensive monitoring combined with simple rule-based automated control of electrical devices resulting in reduced energy wastage - sometimes to the tune of 20% on their energy bills. Solutions have been deployed to

• reduce the dependence on unsustainable energy sources – by reducing unnecessary consumption, improving energy utilization and flattening peaks in consumption.
• increase our dependence on sustainable energy sources -- by exploiting renewables and finding ways to store excess energy from the sun or wind during periods of low consumption.

With the already large share of buildings’ energy usage rising further, the replicability of these solutions to institutional buildings, such as schools, IT-intensive offices, government establishments, etc., make these solutions attractive.

Basic rule driven control, for example, specifying when some appliance should be turned ON or OFF, can in itself result in quick Return on Investment (RoI). More complex control logic, such as controlling multiple HVAC units serving a common space or varying the AC load based on footfall in one part of a building, is also possible.  

Data collected during the operation of the energy manager can be analyzed and mined to improve upon the rules, to deliver even better performance.

Figure below shows six different patterns of consumption in one apartment complex,  with 60 households. A peak in the consumption of an apartment  might occur when many energy-consuming appliances are ON at the same time in that apartment. For example, in the 5th graph we notice two distinct peaks,  one around 9 am -- perhaps because people are taking hot water showers before leaving for work, and another around 9 pm -- perhaps because food is being prepared while the TV is ON.)

In addition to the increasing gross demand, peak demand has also been rising, causing concerns over increasing costs, poor quality of power and the depletion of resources; the continued dependency on fossil sources will have a detrimental impact on the environment. Furthermore, peak demands are handled by turning ON stand-by power generators that add to the capital costs,  increasing per-unit generation cost. About 20% of the generating capacity exists in a power grid to meet the peak demand, which occurs less than 5% of the time. The quick-responding oil/gas fired (highly polluting) generating sources exacerbate damage to the environment. The latter can be seen in the form of dramatic climate change and the alarming increase in the number of health problems due to the ever increasing environmental impact of the huge demands for more energy.    Since a smoother load profile improves grid stability and quality of service, flattening of peak demand is an important energy-challenge, requiring  suitable demand-response (D-R) techniques. In general, Demand-Response techniques are designed to:

• Avoid / reduce consumption (e.g., turning off devices when not needed). Simple motion sensors are often deployed in places like restrooms to turn off lights when they are unoccupied.
• Optimize/ balance demand and supply (e.g., by setting optimal comfort levels and scheduling appliances).  For example, most offices see increased energy usage between 2pm and 4pm, which can be mitigated by pre-cooling; energy needs of homes can be managed by better scheduling of appliances such as washing machines and dishwashers.   
• Flatten  peaks in consumption profile (e.g., by shifting time of operation). Research on flattening or reduction in peak demand spans over the gamut of work from minimizing peak demand through buffering of energy from renewable resources to the use of predictive control  in building HVAC systems.
• Store excess energy (e.g., in batteries, which can also help handle blackouts). Energy Storage Systems (ESS) aid in smoothing out this cyclical and stochastic power flow thus complementing distributed generation systems. Under time-varying pricing schemes, ESSs also reliably allow shifting power consumption to low price periods thus cutting down the electricity bill. The maximization of this economic benefit is achieved by smart scheduling of the ESS under uncertainty.
• Exploit renewable sources (e.g., rooftop PV, wind)   Rooftop photovoltaic systems allows buildings to reduce grid power dependence by harvesting solar energy. For example, a recent study in Mumbai  concluded that the total rooftop PV installation potential for Greater Mumbai is around 1.72 GWp which if fully harnessed can reduce the dependence on the grid almost by half.

These considerations have led to the proliferation of Building energy Management Systems (BMSs). A smart BMS should also be able to perform tasks like reducing and optimizing power consumption, monitoring the status and health of the `appliances in the building, maintaining expected energy consumption in different parts of the building, profile energy consumption of different areas, identifying zones with anomalous power consumption, to name a few.   A BMS delivers these services by tracking, using sensors, various pieces of information like environmental parameters (temperature, humidity, etc.), occupancy status and count, energy available and cost of control. A  BMS  can also flatten the power consumption peaks seen in buildings. It can employ other techniques, such as use of renewables, use of storage devices for storing excess energy during times of low consumption, etc., which can be brought to bear on the energy problem

With many start-ups in this space, making immediate use of the low-hanging solutions, there is a real excitement in the air. Business models including SaaS (Software as a Service), whereby the hardware cost is borne by the solution provider and monthly fee is charged, benefit users from day one. Other innovative models can also serve as attractive ways to spur wide scale adoption of the many solutions.  

But, conservation of traditional energy sources while diversifying into modern renewable energy sources and integrating them into the grid  offers many challenges including:

• Stochasticity of renewable sources introduces significant technical problems for their integration into the existing power grid. Vagaries introduced by changes in cloud cover and wind speed imply increased unpredictability in the load imposed on the electric grid, complicating the task of scheduling power generation.

• Energy has become a commodity, with wholesale markets seeing dynamic real-time prices. Retail markets  typically provide flat rate contracts to end-consumers but even there time-of-use pricing is not far off. So it makes sense to provide incentives to users to shift loads from high-price hours to low-price hours of the wholesale market. The decisions also depend on the availability of different sources of energy.
• In the presence of a large number of sensors, enormous amount of data is generated. The data may have issues like missing values, corrupted values, and inconsistencies. These can further complicate the process of energy management and also introduce other problems, e.g. privacy.

• Localized heating and cooling systems (that often use community waste as heat sources), commonly found in EU countries, may be applicable elsewhere as well. But capital costs may deter their adoption.

Fortunately, as evidenced by the recently concluded ACM eEnergy Conference in Hong Kong, researchers from around the world are busy addressing these and other problems, so one can expect an even better set of  energy management tools and techniques to be available in the coming years.

When people become more aware of the energy problem, we can expect them to take the necessary steps to manage energy more effectively in their offices and in their own homes. But, experience suggests that there is a  need for smart energy management systems which can enable people to tackle their energy needs without manual intervention or without changing their behavioural pattern -- with  users’ support needed just to install the automated solutions. That will lead to a millions-fold multiplicative reduction in energy consumption, helping realize energy savings that  will cross the tipping point.

This indeed is the holy grail!