New Issues Looming Over The Horizon

Written by Sioe T. Mak

Can a system designed for automatic meter reading be economically expanded or upgraded to implement future added-value capabilities without requiring a major overhaul, a large increase in capital expenditures and future added utilization costs? The question is timely and urgent, as smart metering gradually becomes ubiquitous.

Researchers are exploring a wide variety of smart applications to optimize and improve the reliability of the distribution network. Demand response, outage management, assets management, distribution automation, and novel rate designs are among the possible features that could be part of the smart grid. But, it is still too early to assess which way the industry is moving in terms of priorities, primarily because of uncertainties about the capability of the existing technologies to enable implementation of the various new functions.

Prospective communications technology provides a good example of why the industry is so hesitant. The rollout of broadband power line technology has been a fiasco. To understand why this has been such a rocky road, it helps to recall just how complicated the electrical distribution environment is.

The utility distribution network typically covers large geographical areas. The physical energy delivery infrastructure is designed to operate at 50 Hz or 60 Hz AC at voltage levels ranging between 120 and 480 Volts at the customer level and 4 kV and 34.5 kV at the medium voltage level. Distribution transformers link the medium and service voltage levels. Capacitor banks are dispersed at the medium voltage circuits to improve the system power factor and improve the voltage profiles at the customers' premises. The distribution network circuit can consist of overhead wires or underground cables, single-phase or three-phase with or without ground wire.

Taking those basic features into account, communication engineers have been looking at the possibility of using the electrical distribution system to provide voice, Internet and streaming video services. But this literally implies a new and novel broadband powerline communication capability operating at carrier frequencies that are several orders of magnitudes of the 60 Hz power frequency. As it turns out, effects due to multiple carrier wave reflections from junctions, attenuation due to the capacitor banks, cross-talk between phase wires, signal transfer difficulties across the distribution transformers, the need to install many repeaters and signal transfer devices across transformers make system design extremely complex, unreliable and uneconomical.

Multi-hop low power RF technologies have problems of their own in downtown areas of large cities and in rural areas. Too many nodes are required in rural areas where very few customers are served. In metropolitan areas, high rises pose problems for line-of-sight requirements for signal propagation. During the initial implementation, nodes literally follow the streets in the downtown areas, but to access the meters in the basements of the high-rises a different technology is needed.

Against that sobering backdrop, new concerns are looming over the horizon. Electric and hybrid electric vehicles are slowly coming down in price and eventually will be within reach of the general population. Most likely they will need to be charged every night for fairly long periods of time. Unlike air conditioners or water heaters, which are fairly short-duration cyclic-type loads, charging of vehicles batteries does not lend itself easily to demand shifting applications aiming at reduction of peak usage.

In addition, there are harmonics to contend with. Among the many problems that need to be resolved are: voltage distortion, additional losses in the network, increase of third harmonics currents in the neutral, circulating third harmonic currents in the delta connected windings of 3-phase delta-wye connected step-down transformers, and voltage and current distortions affecting energy metering accuracy.

Rooftop hybrid wind-photovoltaic systems with capacities ranging between 2.5 kW and 5 kW are coming down in price, making them more affordable to the relatively more affluent. Operating at low power they are connected directly to the service voltage network at the customer premises. But with sufficiently large numbers of such wind-solar systems scattered throughout the distribution network, a short circuit at the distribution network, leading to post-fault islanding phenomena, creates unique problems.

When a fault occurs at the feeder serving a group of customers, the total fault current at the fault is the sum of the fault current contributed by the distribution substation bus and the total sum of all the fault current contributions of the wind-power systems beyond the fault site on the feeder. The contribution by the individual system to the fault current could be of the order of magnitude of the maximum household load current depending on the impedance between it and the fault location. However, this current flows in reverse direction through the utility meter into the utility network; the effects on energy metering at the customer house are unclear.

After the fault is cleared there is the possibility that islanding conditions prevail. If the wind-power systems are not disconnected from the network, all the units will try to serve the loads as part of the disconnected network, which may be larger than the total capacity of all systems. Loss of synchronism will create power swings among them at frequencies that are as yet to be determined. The customer meters at homes where the system is installed will see the energy flow directional changes; it is still not clear how meters and energy storage devices at the customer premises would be affected.

The time has arrived to start investigations into what near-term and long-term impacts these new issues will have on present attempts to install the more familiar smart grid applications at the distribution network. Demand response applications may have to be modified if a large number of electric vehicles are served by the utility because of the long charging times of the batteries plus the harmonics that these chargers generate. Asset management applications such as load and voltage balancing to reduce line losses, integrated Volt-Var control to improve voltage profiles and system power factor, and outage management damage and safety issues are real engineering problems that electric utilities have to resolve and of which the industry may yet not be aware.



s mak

Sioe T. Mak, an IEEE life fellow, is an associate consultant with ESTA International, LLC. Previously he was senior staff scientist at Distribution Control Systems, Inc., a subsidiary of ESCO Technologies Corp. He has served on many IEEE committees, published widely in areas such as power frequency communication technology and smart grid applications, and holds numerous U.S. and world-wide patents. His diploma in electrical engineering is from the University of Indonesia, and he earned an M. Sc. and Ph.D. in electrical engineering at the Illinois Institute of Technology in Chicago.