Digital switch

Obsolescence, ageing, performance issues, and reliability are driving the US nuclear power industry to upgrade its instrumentation and control (I&C) systems and take advantage of digital technologies. As analogue equipment is no longer made, the industry has had no choice but to convert to digital I&C not only for its availability from a number of viable suppliers, but also for improvements that it can make to plant reliability, efficiency, and safety. For example, the Oconee Nuclear Station, which has three Babcock and Wilcox (B&W) pressurized water reactor (PWRs) units has upgraded two of them and is planning to complete I&C upgrades in the other unit in the autumn of 2013 (see also NEI November 2010, pp12-6). Though Oconee does not represent the first digital I&C upgrade in the US, it is the most important one to date as its upgrade included the reactor protection system (RPS), engineered safety (ES) system, and nuclear instrumentation system (NIS). To accomplish this upgrade, Oconee used the TELEPERM XS technology from AREVA.

The total cost of the upgrade for the three units is said to be about $300 million (that is, approximately $100 million per unit), including equipment, vendor engineering, in-house work and installation. This is, of course, significantly higher than originally estimated. Moreover, much of it is attributed to the delays caused by regulatory issues that had to be resolved insofar as Oconee was the first US nuclear power plant to convert its RPS, ES, and NIS to digital. In fact, much of this $300 million is said to be for the first unit; however, the $100 million figure for each unit jibes with other rough estimates that the author reached for a digital upgrade of a plant protection system. To reach the estimate of $300 million for the total Oconee I&C bill, the author used a $250 million figure that had appeared in local and national news reports as well his own assessment based on potential cost overruns and other unanticipated costs.

Cost

The costs of digital upgrades that have occurred in the worldwide nuclear power industry to date average about $100 million per unit, based on an informal survey of the nuclear industry carried out by the author. In each of these upgrades, the author assumed that the scope of work included the digital I&C hardware and software for RPS, ES, and NIS, documentation including quality assurance (QA) and software verification and validation (V&V), and a reasonable level of engineering work and regulatory interactions.

Some have estimated digital I&C upgrades including RPS, ES, and NIS as low as $25 million per plant unit. However, the author believes these estimates are unrealistic; they would only be achievable if a utility with a large fleet (for example, EDF’s French fleet or Exelon’s US fleet) bought the products all at once for its whole fleet from the same supplier and performed much of the engineering, installation, and regulatory work on its own. Of course, to estimate a cost of a digital upgrade, one must know the scope of the work, the procurement details, and the utility requirements. Furthermore, the cost would very much depend on whether the upgrade is the first one in the fleet or one of many.

Market

In Table 1, the potential price of a digital system upgrade in a US plant is estimated, along with an indication of how the cost might decrease once the first unit in a site or in a fleet has completed its upgrade. The cost is assumed to be high for the first upgrade, and to decrease significantly for additional units in the same site or same fleet. To arrive at the numbers presented, the author spoke with experts across the industry who are familiar with the subject. The consensus was, that although any estimate of the current value of the I&C upgrade market in the US is difficult to develop, there is great interest in the industry for at least a rough order-of-magnitude idea of the potential upgrade market and costs.

The cost figures in Table 1 are meant to imply the price that a utility would pay a vendor for the product, including the hardware, software, engineering and paperwork. Also shown in Table 1 are the probabilities that a US plant would upgrade over the next ten years, and the total market in the US for each upgrade. We have shown only 11 systems, although there are other systems in the US fleet that may be upgraded. Based on this information, the total US market for I&C upgrades of main equipment is about $5.4 billion, accruing to the I&C vendors. Assuming that 55 US plants may upgrade over the next ten years, the per-unit cost is about $100 million. Given that the value of an old 1000-megawatt plant is about $4 billion, this represents 2.5% of the cost of the plant for a huge improvement in plant reliability, safety, and efficiency.

Note that in Table 1, the probability of an RPS upgrade in US is estimated to be about 50 percent, while that of an NSSS control system upgrade could be as high as 80 percent. This is because RPS system upgrades do not normally add much to the bottom line; while control system upgrades can improve reliability, reduce plant trips, and thereby add to the plant profitability, enticing the utilities to spend the money. Today, most RPS systems in the US fleet are working reasonably well, although they are old and obsolete.

In Europe and Asia, the move from analogue to digital I&C has occurred at a much higher pace than the US. Even today, after three decades since digital I&C systems were introduced to the US market, only about 20% of plants have converted from analogue to digital; these are mostly in balance of plant equipment, and almost all for non-safety applications.

Diablo Canyon is said to be the next plant after Oconee to implement digital I&C upgrades for at least a part of its RPS. This job is expected to be a pilot project in the 2012-2014 timeframe and to be completed by or before 2015. The digital equipment for Diablo Canyon will be supplied by Westinghouse and Triconex (Invensys), and the cost estimates for the first phase of this project is about $25 million. The consensus is that the floodgates for digital I&C upgrades should open after Diablo Canyon’s successful implementation, and the time estimate for the start of the first wave of upgrades is probably in the mid-2015.

Regulatory issues

Regulatory questions about defence-in-depth and diversity (also known as D3), and potential errors in operating systems or software used in redundant trains (also known as common-mode failure) are among the hurdles that the US nuclear industry is facing as it contemplates implementing digital upgrades for important and large systems, especially those that can affect safety. Unfortunately, there are enough concerns involving digital equipment to keep the US regulators worrying about digital I&C for safety systems. That said, the nuclear industry outside the US has proved that digital systems are very safe and can add significantly to the efficiency and reliability of nuclear plants.

This is in addition to the excellent experience of the aviation and aerospace industry with digital equipment in all facets of their operations. Today, passenger planes are operated with fully-digital control systems with few or no mishaps that can be attributed to digital technology. For example, the tragic crash of the Air France flight from Rio de Janeiro, Brazil to Paris, France in 2009 was first feared by some to be due to potential problems in the plane’s automated control system. It was later traced to malfunctioning pitot tubes, which are used to measure the speed of the aircraft.

Suppliers of digital I&C

Although the market for digital I&C systems for safety related equipment in US nuclear power plants has not developed rapidly (especially in the US), numerous manufacturers of digital I&C equipment are ready to serve the nuclear industry. Table 2 provides a list of most of these suppliers. In the US, however, the main contenders are:

Safety systems

• AREVA
• Invensys/Triconex
• Westinghouse

Non-safety systems

• Westinghouse
• Invensys/Triconex
• GE M&CS
• Siemens

Others, such as Rolls-Royce, Lockheed Martin, and Northrop Grumman, are also striving to serve the US market and are expected to be able to capture a portion of the I&C upgrade market. In 2011 Rolls-Royce was awarded a contract for $325 million to provide new I&C equipment to much of the French fleet of PWR reactors. The Rolls-Royce offering will include the Spinline system for RPS, new digital NIS, and digital rod control system. To add to the capabilities of new plant equipment, the author’s organization (AMS) has recently developed new online condition monitoring (OLM) to verify the performance of I&C systems, in-site cable condition monitoring for aging management, diagnostics technology for rod control systems of PWRs as well as NIS cable/connector testing using remote in-situ technologies. It is expected that the next generation of I&C equipment will not only incorporate digital technology but also have equipment with advanced diagnostics and automated predictive maintenance features.

The digital I&C suppliers have in recent years integrated field programmable gate array (FPGA) technology into their digital I&C offerings (NEI June 2012, pp24-6). This development pleases the regulators because FPGA-based products are less susceptible to software errors. With FPGA, many of the commands are executed in the hardware, reducing regulators’ concern with software errors and common cause failures (when a software error propagates throughout a system).

Digital I&C equipment will inevitably find its way into the US nuclear fleet, not only for plant controls but also in plant protection systems.

This article first appeared in the November 2012 issue of Nuclear Engineering International magazine.