AC vs. DC Power A New Battle of the Currents

Spread knowledge

Advances in technology and the increase of devices that can generate or operate on DC power are again raising the debate over AC vs. DC power. Eliminating conversions from DC to AC and vice versa can greatly improve the efficiency of both the grid and the devices that can operate in a DC mode. Solar photovoltaic panels, batteries and fuel cells generate/store DC power, and residential, commercial, and industrial facilities are projected to increase DC electrical loads that do not require first converting to AC.

In developed regions of the world where the AC power grid is well established, it may not be feasible to change over large portions of the existing grid to DC. However, certain aspects of the grid, such as distribution primaries and secondaries may be candidates for a DC revolution. World regions where electric infrastructure is developing may have a greater opportunity to take advantage of the economic benefits of being able to adopt the latest innovations in DC energy production, transportation, and end-use technologies.

With this in mind, several IEEE Smart Grid experts were asked to provide their input to the question:

What do you see as the future of AC and DC power, relating to various world regions?”

Doug Houseman, an IEEE member active in the IEEE Power & Energy Society’s Intelligent Grid Coordinating Committee, is vice president for technical innovation at EnerNex

It is not so much by region but by access to existing infrastructure. In North America today there are almost one million homes and businesses off the grid. Most of them run on DC systems. Thanks to the Recreational Vehicle Industry (RV) there is a robust set of appliances and other household items that are designed to either run on 12- or 24-volt DC. While most do not meet the Energy Star rating and they also are a little more expensive (and typically smaller in size) than standard appliances, they are readily available.

In the developing world where renewables will be the primary source of electricity, the vast majority of renewables are DC-based, as are the batteries that will store that power, so expect that village systems in developing areas will all be DC-based. The market for high-quality DC-based systems will grow rapidly over the next decade, as standards are created that allow anyone’s equipment to work with anyone else’s equipment and the cost of generation (e.g. Photovoltaic cells) continues to drop.

In the developed world, large amounts of energy delivered as AC is now consumed as DC. In computers, consumer electronics and many small appliances as well as LED lighting the actual power consumed is DC. That means there is a conversion loss that adds to the energy usage. Similarly there is energy loss in converting from the DC produced by the photovoltaic systems on the roof to AC and then in many cases back to DC. Add electric vehicles and local storage, and the conversion losses rapidly mount.

It may be that the right answer is to use a mixed delivery system in homes and businesses moving forward. To this end the IEEE Power and Energy Society has started a project to look at the use of DC in the home. This project includes a business case, use cases, a standards review, a building code review and development of a list of research that should be undertaken. This project will kick off at the IEEE PES General Meeting in Vancouver in July.

Sam Sciacca, IEEE senior member and president, SCS Consulting, LLC

Telsa or Edison…….Who was right? Turns out they both were!

Traditional AC power will continue to be the primary mechanism for the majority of transmission and distribution networks. This is largely due to fact that it is far easier to convert AC from one voltage level to another, which is required for moving bulk power across the grid efficiently and reducing losses. AC power is also easier to control/interrupt in switching and fault situations than DC, making the capital investment to build AC infrastructure far less than DC infrastructure. DC transmission systems have been in place for many years, but the experience that has been developed with the technology has relegated DC transmission to limited and highly specific instances.

Where DC holds a tremendous amount of promise is in the home/office/building environment where local generation (wind or photovoltaic) is available. The conversion of photovoltaic (DC) power to AC, only to have that AC power converted back to DC for many home electrical devices, is an incredible waste of energy. Eliminating this waste, by some estimates, could improve photovoltaic system performance by as much as 25%. Additionally, wind turbines are both more reliable and less costly in DC configurations, due to the greatly reduced complexity of the mechanical transmissions that are required for turbine AC generation. However, there are still some technical and economic obstacles that need to be overcome.

1.   Utility interface – in order to connect the DC-capable home to the utility distribution system, a new type of device is needed. This new device, called an energy router, would be able to take the DC from local generation and convert it into AC, and convert the AC distribution voltage to DC for home supply. Such devices are currently in R&D, and need further development to determine the practicality and lifecycle costs.

2.  The majority of building codes around the world do not cover DC distribution in homes, offices, and commercial buildings. Wiring practices, distribution panel ratings, plug/socket conventions, grounding practices, circuit breaker devices, all will need revamping to accommodate a DC option.

The above are currently the topics of two IEEE-SA Industry Connections activities. (More information can be found at

Playing a major part in the equation is the worldwide interest in energy savings, recycling and the appearance of new products that can operate on DC. As example, the elimination of cathode ray tube displays (CRT) in televisions and computer monitors has eliminated the need for AC in these devices, which was much more efficient than DC for CRT operation. New LED lighting must have DC to function, either converting AC to DC at the light fixture itself or from a DC power circuit. And the presence of an electric vehicle with a charging system in the home greatly furthers the attractiveness of having DC power available, even permitting the car batteries to be used as back-up power to critical circuits in the home without costly inverter technology. However, even if DC were made available in residential/commercial environments, there are still devices/appliances (air conditioners, refrigerators, freezers, well pumps), which are much more efficiently built/operated/maintained in an AC configuration. Given that these are sometimes the predominate loads of a home/office, it may be difficult to avoid not building both AC and DC distribution circuits in the structure, at a greater cost than building only one or the other.

I expect the situation to continue to evolve as technology, policy and economics conspire to improve the cost-effectiveness of DC in residential and commercial environments.

Nicholas Abi-Samra served as the General Chair and Technical Program Coordinator for the IEEE General Meeting of 2012 and Chair of IEEE Power & Energy and Power Electronics in San Diego. He is senior vice president, Electricity Transmission and Distribution for DNV KEMA Energy & Sustainability. 

The grid as we know it today is basically an AC grid, with a few, disproportional number of High Voltage Direct Current (HVDC) systems. It has worked well as the backbone of the transmission grid for over one hundred years, and will continue to do so for the foreseeable future. There is too much invested in it to think otherwise. HVDC has been used to connect regions with different frequencies, or when regions have the same frequency but they are not synchronized or for shipping bulk power over long distances. Therefore, the use of HVDC was limited to isolated, special purposes, where AC fell short, for one reason or another. But to put in this in perspective, though the amount of power being handled today by HVDC systems is an order of magnitude more than it was forty years ago, standing at about 60-70 GW, (see chart) this is still less than 2% of the installed global generation capacity.

More usage of HVDC systems may be attributed to HVDC’s inherent properties that make it more convenient and efficient for transmitting power from offshore wind farms or remotely located renewable energy resources on land compared to AC systems.  Even wider use could be possible if certain enablers come into realization. These include:  less expensive and more efficient converters, higher voltage HVDC cables, advancements in HVDC circuit breakers and advances in monitoring and control systems for multi-terminal HVDC systems. One more hurdle needs to be overcome, and that relates to the fact that at present many utilities have a good history of operating the AC system and also utilizing a DC link as an interconnector; however, very few operators have the experience of operating an integrated HVDC link in conjunction with the AC system and possessing the practical know-how of dealing with issues associated with possible overloads on the AC system, i.e., rotor angle instability on the exporting side of a boundary following a fault, etc.

In short, AC will remain as the backbone of the grid for many years to come, HVDC system usage will increase, with the most pronounced growth in regions of large amounts of remote renewable generation (on land or off-shore).  HVDC systems may play a major role also in supergrids, such as the European supergrid concept where the present AC transmission grid almost reached its transport capacity and large transport of electricity through the grid leads may lead to stability problems.

John McDonald is an IEEE Fellow, past president of the IEEE Power & Energy Society (PES) and past chair of the IEEE PES Substations Committee. He is director of technical strategy and policy development at GE Energy’s Digital Energy business.

Historically, DC transmission has proven cost effective only over distances of more than 500 miles, but as new technology is developed, we may see that breakeven point be reduced. With modern power conversion electronics, DC power could capture as much as half the grid within a few decades.
According to the Energy Information Agency (EIA), the fastest growing portion of residential electricity use is consumer electronics and small appliances. These devices primarily run on DC power. However, converting AC to DC power involves wasted energy. Many of these devices have a conversion efficiency of no better than 80 percent and some low-end devices have efficiencies as low as 65 percent in converting power. One estimate says 5 percent of all electricity used in the typical US home is lost to conversion of AC to DC power to run DC devices. A solar photovoltaic (PV) system on your home may lose close to a quarter of its output at the inverter. The same is true for any battery used to store solar output or provide backup power in the case of an outage. Charging an electric car may lose as much as 24 percent converting AC power to the DC power stored in the vehicle’s on-board batteries. This translates into costing close to $100 per year if you drive your EV about 35 miles a day, about the distance of the average American’s round-trip daily commute. Reducing AC-DC conversion losses also reduces the waste heat generated by all the power converters, which further saves on the large amounts of energy used to cool hot-running data centers.

United States
In downtown San Francisco hundreds of customers still buy DC power from PG&E to drive their winding-drum elevators. PG&E has invested a substantial amount of money and technology to upgrade the safety and reliability of its DC grid and better integrate it with the surrounding AC system. PG&E hosts in its service area one of the fewer than two-dozen high-voltage DC transmission lines in the US. The Trans-Bay Cable is a 53-mile, 200-kilovolt DC line that carries power under the Bay from Pittsburg to San Francisco. Two other DC lines, the 500-kilovolt Pacific DC Intertie and the 500-kilovolt Path 27 from Utah, serve Southern California. The Emerge Alliance, which advocates for DC power distribution in commercial buildings, is promoting a new 380-volt DC standard for data centers and commercial buildings. A few years ago researchers at Lawrence Berkeley National Laboratory showed that energy-intensive computer data centers could achieve power savings of about 10 percent by feeding server racks with DC power rather than AC. Intel has valued annual power savings for a medium-sized data center in the U.S. at $1.2 million, and the value should be considerably more in Japan and Europe, where power prices are higher. Emerge Alliance has established a standard for 24-volt DC ceiling circuits and says that running LED ceiling lights on DC lines uses up to 15 percent less energy than doing the AC-to-DC switch inside the fixtures. Emerge is now working on bringing DC power to employees’ desktops, letting them plug in computers or phones without the need for hot-running converter boxes. The US military and large campuses such as universities and hospitals are experimenting with DC systems, particularly to run critical systems in the case of a failure of the AC power grid.

At very high voltages, DC has even lower losses than AC, making it ideal for very long transmission lines – such as those carrying renewable wind power from the North Sea to continental Europe. A DC grid would allow wind to supply at least 30% of the power needed in Europe. Moreover, it could do so reliably—and that means wind power could be used for base-load power supply. A group of Norwegian companies have already started building high-voltage DC lines between Scandinavia, the Netherlands and Germany, though these are intended as much to sell the country’s power as to accumulate other people’s. Airtricity, an Irish wind-power company, proposes what it calls a Supergrid to link offshore wind farms in the Atlantic ocean and the Irish, North and Baltic seas with customers throughout northern Europe.

One market survey predicts that China will use DC for up to 40 percent of its new transmission lines. Construction started last year on the world’s largest Ultra High Voltage Direct Current (UHVDC) power line, and China started on a second 750kV HVDC one on the same day. The 800 kV system will connect Hami prefecture in eastern Xinjiang with the central city of Zhengzhou, along a 2,210-km-long, $US $3.7B power line and is designed to handle 8 million kW of power. The State Grid Corporation of China said they can reduce 317,000 tonnes of sulphur dioxide and 267,000 tonnes of nitrogen oxide which would otherwise be produced during the transportation, a signal that energy losses are finally being recognized.

Some electronics-heavy facilities are now developing all-DC “microgrids” to feed power to users. There are plans for a DC microgrid at China’s Xiamen University. A self-contained electrical grid will span three campus buildings, linking a 150-kilowatt rooftop solar array to LED lighting systems and banks of computer servers.

Since all servers run internally on DC power, the incoming AC power must be converted. Instead of having power converters on each computer, some companies are installing large centralized converters and distributing 380-volt DC power across their server farms. Making that switch and eliminating AC-DC converters on battery backup systems cut power consumption by 15 percent compared with conventional AC configurations. Japan’s telecommunications giant, NTT, has at least five facilities in Tokyo using DC power. Last year it completed a DC-based server center in Atsugi City, southwest of Tokyo, which is its first to serve external clients.

Power Grid Corporation of India Ltd. is building an ultrahigh-voltage direct current (UHVDC) transmission system to supply hydropower from mountainous northeast India to the populous region of Agra in central India, 1,700 kilometers away. Northeast India has abundant hydropower resources scattered over a large area, while the load centers are often located thousands of kilometers away. India plans to create pooling points in the region to collect electricity generated from several hydropower stations and transport it across power superhighways to major urban load centers. The UHVDC link, operating at 800 kilovolts (kV) will have a converter capacity of 8,000 megawatts (MW), the highest ever built. When operating at full capacity, it will have the means to supply electricity to 90 million people based on current figures for average national consumption.

Nearly all of the above factors would seem to favor DC over AC transmission. Two major developments are needed for DC transmission to be technically and economically feasible. Electronic DC/DC voltage transformers, built for high voltage, high power conversion are much more expensive presently than conventional transformers. They must reach cost parity with conventional AC/AC transformers. In addition, DC circuit breakers are a problem, especially at high power levels above one megawatt. One supplier recently announced a breakthrough on HVDC circuit breakers that they say will allow HVDC circuit breakers up to one gigawatt (still well below what will be needed to implement a supergrid). Though not published by the supplier, estimates are the cost will be about 100 times as high as comparable AC circuit breakers.

Large consumers of DC power are already wiring up DC microgrids to more efficiently satisfy their needs. The large data centers that run the Internet, cloud computing services, and telecommunications networks are the best candidates. These server farms consume more than 1.3 percent of all electricity worldwide, and this figure is growing quickly.

The Global Energy Network Institute, based in San Diego, California, says high-voltage DC lines could be used to bring solar energy to market from places such as the Sahara. Wind and geothermal power could be gathered from as far afield as South America and Siberia. Such a globalized market has its attractions.

Carl Imhoff is an IEEE member and manages the Electricity Infrastructure Market Sector within Pacific Northwest National Laboratory’s Energy and Environment Directorate.

I definitely see an increased synergy between AC and DC systems for low voltage end-use loads, particularly at the building-to-grid interface. This is an emerging area of interest because 40% of U.S. energy consumption occurs in buildings and they offer significant potential for efficiency and delivering ancillary services back to the grid. Using locally generated DC power from renewables and distributed generation in DC loads just adds to end use efficiency by avoiding a conversion to AC. Buildings with data centers and renewables on the rooftop realize really good synergy to use as much DC power locally. So I see a really nice synergy between AC/DC at that building grid interface.

There is potential for AC/DC systems in the buildings-to-grid interface in all regions of the world, but it is very interesting in hyper-developing economies such as Asia, where there are doing a lot of new greenfield development. These regions have opportunity to start with AC/DC hybrid systems in new buildings and link them to the grid in ways that support increased efficiency and also better use of the DC coming out of renewable generation!

At the bulk high voltage transmission level – there seems to be a lot more activity going on in Europe and Asia than in North America. HVDC does play an important role in North America in the current transmission system. I think HVDC will expand slowly in the US due to siting and the regulatory environment – it’s very difficult to site new DC transmission assets currently in the U.S.  So, I think we will see more HVDC in the US, but it’s developing slowly. However, there are two emerging opportunities here.

One is the growing interest in multi-terminal HVDC. This has been developed for offshore wind mainly in Europe, but as the costs of multi-terminal HVDC come down, it will make it more attractive for use in the US and that’s a good thing. And if we can solve the regulatory siting challenges, HVDC gets us an interesting degree of freedom for managing reliability in the power system. In addition to expanding long-distance transfer of power, we can also use the DC lines to help manage some of the oscillatory or dynamic constraints in the AC system. So HVDC can help us with reliability as the economics and regulatory processes improve.

I do believe that it’s important to be watching what’s going on elsewhere in the world and be thinking about what is the transmission system we want in 2050 and what role does high voltage DC play in that system. And I think you will see some expanded use of DC in the US. I don’t expect that major expansions of HVDC will be economically viable in the near term in the U.S.  But based upon what we see now with low cost natural gas, it will play some important roles in the long term. Right now I think the big action in HVDC is in Europe and Asia. In the U.S., I see good opportunity for low voltage, DC/AC systems, particularly at the buildings-to-grid interface.

Massoud Amin is a senior member of IEEE, chairman of the IEEE Smart Grid newsletter, and a fellow of ASME. He holds the Honeywell/H.W. Sweatt Chair in Technological Leadership at the University of Minnesota.

If Thomas Edison had had his way in the late 1880s, direct current (DC) and distributed generation would be the norm as all power would have produced and used locally, rendering transmission lines unnecessary and impractical. But Edison was wrong…at that time. Nikola Tesla had a better idea and won the first ‘battle of the currents’ for over a century, thanks to a large part to George Westinghouse who commercialized Tesla’s inventions and built the first power generators at Niagara Falls, to supply markets in New York City.

Today’s global grids operates mainly on alternating current (AC), achieving an efficiency of nearly 97% for High Voltage Alternating Current (HVAC) transmission lines, which constitute the backbone of nearly all power grids.

In Tesla’s prevailing model of electrical generation and distribution, large turbines produce several hundreds of megawatts of electricity and then a transformer steps up the voltage so it can be transmitted long distances with minimum loss. At the other end the voltage is reduced to a lower level by a step-down transformer, so it can be delivered to end-users.

As a result, today’s power grids are networks of generation plants, transformers for stepping voltages up/down, DC/AC and AC/DC converters, high-voltage transmission lines for long-haul transfers, and low-voltage distribution lines and substations for short-haul transfers into homes, offices, and factories. While Edison, supported by J.P. Morgan and his team, is famous for inventing the light bulb, Tesla envisioned and invented the entire infrastructure needed to power those light bulbs.

Fast-forwarding to today, some 120 years later, a new battle of the currents is arising as digital technologies and appliances such as computers, mobile devices, and radios use DC instead of AC. Conversion from AC to DC requires copious use of small transformers familiar to modern consumers who must constantly recharge a variety of devices. For example, from fuel to lighting, end-to-end system efficiency when electrons are generated from coal-fired plants to provide lighting for incandescent light bulbs, the efficiency is a mere 1.6%, resulting in losses that amount to 98.4%. As further illustration, fossil fuel-based electric power generation, which account for 70% of all electricity production in North America, average an overall end-end system efficiency of about 22%.  To improve efficiency, and other increasingly important performance objectives such as power quality, better overall system operation, and more precise control call for increased use of DC technologies to minimize the number of conversions.

From the overall system perspective, the potential for High Voltage Direct Current (HVDC) state-of-the-art transmission technologies has increased. The advantages of HVDC transmission over conventional HVAC technologies are well known for long distance, point-to-point power transfers. Another unique capability of HVDC is its ability to connect asynchronous grids prevalent in national and regional power networks. HVDC technologies provide extremely fast response times, the potential to control power flows, thus faster and more accurate systems operation and control, and the ability to segment parts of the power system quickly (or even automatically as part of a self-healing grid), all of which can enhance the flexibility, reliability, and resiliency of the grid.

Currently, advanced HVDC transmission projects are being proposed, designed and built around the world – especially in Europe and Asia. Thus far, HVDC deployment activity within the US has been limited, although all major transmission studies at the AEP, EPRI and the US DOE call for adding HVDC lines to strengthen the HV backbone for increased reliability, energy security, to enable electrification of transportation and integration of distributed resources including a much higher volume of domestically available sources, including renewables.

New loads and generation sources, such as solar and wind, will often be connected to the distribution grid that has been traditionally designed to uni-directionally distribute the output of the bulk electric system. Accommodating these new loads and generation sources will require both bridging the gap between the transmission and distribution grids and crossing the chasm between federal regulation of the transmission grid and individual state regulation of distribution networks.  The importance of meeting these challenges is heightened because the economic feasibility of many of the new technologies is dependent on their serving both the transmission and distribution grids simultaneously.

Indications of pathways to that future are here now, as technologies and societal needs have made it not only possible but necessary to take a fresh look at judiciously incorporating both AC and DC technologies at appropriate scales of deployment. For example, HVDC transmission systems are based on the rectification of the generated AC and then inversion back to AC at the other end of the transmission line. Modern systems are based on thyristor valves (solid-state power control devices) to perform the AC/DC/AC conversions. Conventional HVDC transmission systems have been built with power transfer capacities of 3000 MW and ± 600 kV. A new class of HVDC converter technology has been introduced in the last few years, referred to as voltage source converters (VSC), and it is based on gate turn-off switching technology or insulated gate bi-polar transistors (IGBT). These devices have higher switching frequency capability.

HVDC transmission is used in long distance bulk power transmission over land, or for long submarine cable crossings. Altogether, there are more than 35 HVDC systems operating or under construction in the world today. The longest HVDC submarine cable system in operation today is the 250 km Baltic Cable between Sweden and Germany. In addition, more deployments of 800KV DC lines in China have been constructed.

To highlight further opportunities where science and technology from other industries could possibly be identified to fill these gaps we must address microgrids – AC and DC, both self-contained, cellular and universal energy systems, and larger building or campus-sized systems.
Advanced (post-silicon) power electronics devices (valves) to be embedded into flexible AC and DC transmission and distribution circuit breakers, short-circuit current limiters and power electronics-based transformers.

In considering the whole North American system, our first strategy should be to expand and strengthen the transmission backbone by adding about 42,000 miles of high-voltages transmission lines to the existing 450,000 miles of 100KV and higher, at a total cost of about $82 billion. Most of these will be HVDC lines. At the other end, they’ll be augmented with highly efficient local microgrids that combine heat, power and storage systems, as a smart grid with self-healing capabilities (total cost, $17-24 billion annually for 20 years). Many of these end-use facilities – campuses, buildings and homes – will be fully DC.

On options and pathways forward, I am often asked “should we have a high-voltage power grid or go for a totally distributed generation, for example with microgrids?” We need both, as the “choice” in the question poses a false dichotomy. It is not a matter of “this OR that” but it is an “AND.”  To elaborate briefly, from an overall energy system’s perspective (with goals of efficiency, eco-friendly, reliability, security and resilience) we need both 1) microgrids (that can be as efficient and self-sufficient as possible, and to island rapidly during emergencies), AND we need 2) a stronger and smarter power grid as a backbone to efficiently integrate intermittent and geographically distributed renewable sources into the overall system.

By the year 2050, although predicting the future is a risky business, based on our strategic analyses incorporating a wide range of complex drivers, accounting for uncertainties and needs, we envision a drastically different electric grid than what exists today. This futuristic vision of grids with efficient markets, idealized grid-pervasive demand-response, rapid real-time end-point control, smart peripheries with fully coordinated networks of microgrids, synergistic electrified transportation, green and automated distribution systems and efficient AC-DC transmission systems is not only achievable but essential for energy security, innovation-enabled economic growth and the next generation smarter, more resilient infrastructure. This system, customized for available resources and needs, will effectively and securely meet demands of a pervasively digital society in the face of extreme events and climate change while ensuring high quality of life and fueling GDP growth.