Edison’s revenge? High-power DC gets another chance

DC-based power transmission lost out to AC, but DC is making a comeback.

Many things we now use as basic knowledge were not fully understood in the early days of electricity and power in the late 1800s/early 1900s. Still, one thing was grasped: ohmic losses (often called IR or resistive losses) in power-transmission lines were unavoidable due to the laws of physics. The only solution was to increase the line voltage used for power transmission and decrease the line current by the same factor.

Since ohmic power losses are proportional to the square of the current (P = I²R), the savings are exponential and significant. Novice electrical engineers today quickly learn this basic lesson about the relative efficiency of higher voltages for power transmission early and often, even on smaller-scale projects such as PC boards and chassis.

The big issue of those early days of electricity, motors, and lighting (there were no electronics yet) was whether to use AC or DC power for the generators and transmission lines. In those early days, there was a highly publicized battle between Edison, on one side, as a staunch DC proponent, and Tesla, Westinghouse, and others who favored AC. Power transmission using AC could be easily stepped up/down via transformers as needed and thus support longer transmission-line distances, while DC was limited to shorter runs and needed many relatively local “neighborhood” generators to serve neighborhoods; this made some sense as electricity was first adopted for dense urban areas.

AC eventually prevailed for these and other reasons, and we now have 50/60-Hz power distribution and systems worldwide. High-voltage AC (also known as HVAC, but not to be confused with heating, ventilation, and air conditioning systems) came to dominate. At the same time, high-voltage DC (HVDC) was relegated to a few niche situations.

In principle, the arrangement of an HVAC system is straightforward: generate the power from an AC source, step it up via one or a series of passive transformers to tens or even hundreds of kilovolts for transmission, and then step it down again via transformers to the 120/240 VAC needed for the end user. In contrast, the step-up/step-down components needed for HVDC are active and more complicated (Figure 1) and were not practical or cost-effective until a few decades ago.

Figure 1. The HVDC power path involves many conversion stages on both the source and load side, which complicates the implementation challenges. (Image: Duke American via Texas Instruments)

Moving ahead about 100 years to the recent past, the situation has changed dramatically. Many transmission-line designs are using HVDC, and many new HVDC installations are also being planned. There are many reasons, but among them is the familiar, mutually beneficial, positive-feedback push-pull of applications and technology advances: new components make new approaches increasingly viable while that increased viability increases design-ins and the demand for these components, which in turn increases demand for new components, and the cycle begins again.

In the case of HVDC, advances in technologies such as high-voltage/power IGBTs and thyristors have been among the many new components needed (interestingly, many of these same devices and technology were initially developed for all-electric and diesel-electric locomotives). For HVDC, the most common types of converter stations are line-commutated converters (LCCs) and voltage-source converters (VSCs). There’s lots of exciting design work for LCCS and VSCs and related HVDC subsystems using new components and unique design rules.

Other AC versus DC subtleties

In addition to the apparent differences between AC and C, two interesting ones are worth citing. AC and DC currents are dangerous, but AC is about five times more dangerous to those who touch the energized line (don’t ask how this has been determined).

The cyclic voltage changes in AC (50 or 60 times per second) are the main reason behind this enhanced danger. Muscular contractions are more common with AC, and AC also stimulates more sweating. This process lowers the skin’s resistance and makes it more susceptible to electrical damage. The relative danger of AC in lower-voltage consumer users was one of the many points Edison publicized as he made the case for DC. Of course, the risk is extreme at the higher voltages associated with non-consumer higher-voltage transmission lines, and the difference in risk is insignificant.

Another difference is well documented and understood. When a high-power (current and voltage) contact opens, an electrical arc will form between the opening contacts when sufficient voltage and current exist in the system. The arc persists until the voltage required to sustain it is greater than the voltage supplied by the circuit. These arcs are detrimental as they corrode the contacts and, if large enough, can fuse them, which contradicts the goal of an open contact pair.

There is a major advantage to arc interruption with AC power; the voltage will periodically drop to zero and force the arc to extinguish, and at 60 Hz, this limits the maximum arc time to less than 8 milliseconds. A practical scenario is that most AC switches operating under 300 V will usually extinguish the arc as the circuit voltage passes zero. This vastly simplifies the design of switches within this voltage window.

DC circuits do not have the zero crossing, so switches for them require more careful engineering. Several design modifications help achieve more rapid arc extinguishment, but they add complexity to the contact design.

The market changes

How is HVDC doing? The short answer is “very well,” but many numbers depend on who and what you ask. Different market research organizations each have their high-precision numbers, looking out to five years and more (how they can predict the future to three and even four significant digits always fascinates me). It varies depending on what they include in their HVDC market assessment and their methodology. Still, the rough consensus is that the HVDC market ranges were around $10 billion for 2020 and 2021 each and is expected to grow to between $17 and 18 billion by 2026, for a compound annual growth rate (CAGR) of between 6 and 8%. In contrast, the all-important CACG numbers for HVAC growth are several points and billions of dollars lower.

With respect to HVAC versus HVDC costs, a rough guide is that if you exclude the cost of conversion circuitry at each end – and that’s a very big “if” since it is admittedly a large cost – HVDC lines make sense for distances greater than around 500 miles/1000 km for overhead lines, between 15 to 30 miles/30 to 60 km for submarine cables, and 30 to 60 miles/60 to 120 km for underground cables. While conversion circuitry for HVDC is more expensive, that is balanced out somewhat as its transmission lines require smaller towers and, in their simplest arrangement, need only two conductors rather than three. Determining that total-cost breakeven point takes fairly sophisticated analysis (Figure 2).

Figure 2. Determining the crossover point in HVAC versus HVDC economics requires extensive end-point terminal and line cost analysis. (Image: Electrical4U)

While cost is a major consideration, the HVAC versus HVDC case also has important technical aspects. First, HVDC is asynchronous and does not require synchronization among the many sources to maintain stability, so these can be added/dropped as needed or available; in contrast, AC sources require careful initial synchronization procedures and must not be allowed to go out of sync. Further, the power factor of the DC line is always unity, so no reactive compensation is needed. Also, there is no “skin effect” as with AC lines, a detrimental aspect that reduces the effective current-carrying cross-section of the line and increases resistive loss.

In short, there’s a good case for using HVDC. The case gets stronger as the grid increasingly counts on a multiplicity of distant sources such as wind power, solar farms, and even battery-based energy storage systems rather than generation plants. We’re even seeing some commercial power systems use 350 VDC as the primary power line for office buildings, factories, and large apartment complexes, with conversion to 120/240 VAC for the end applications.

There’s an ironic bookend to Edison and DC’s resurgence story. While Edison’s DC is seeing substantial growth, the descendant of the company founded to promote it has gone the other way. Industrial giant and household name General Electric was formed in 1892 through the merger of Edison General Electric Company (which Edison founded through various specialized companies he started) and Thomson-Houston Electric Company.

However, by the first decade of the 21^st century, over one hundred years later, General Electric fell on hard times for various reasons. In 2021, the company announced it would divide itself into three new, totally independent public companies: GE Aerospace, GE HealthCare, and GE Vernova, respectively, focused on aerospace, healthcare, and energy (renewable energy, power, and digital). In short, mighty GE has disappeared.

Even if HVDC takes on a larger role in power-line transmission, Edison won’t get a full victory. HVAC and HVDC each play a legitimate role and will be used where they make the most technical and economic sense. Perhaps the most we can expect is that in their afterlife, Edison and Tesla will recognize that reality and at least become “frenemies?”