When I left off my tale back in late June, I had found out about the resistance to new ideas from the people building and operating light water reactors and I was curiously picking through various technical papers to find out what had been done to develop closed cycle gas turbines (CCGT) to solve real problems.
Energy dense, no pollution, low cost fuelWith detailed knowledge of light water reactor heated steam plants and a good academic and practical knowledge of simple cycle combustion gas turbines, I was enthusiastic about the potential of combining the best of both technologies. The pollution free, high energy density fuel that made it possible to power a submerged submarine for months at a time between maintenance and crew rest periods and for a decade or more between fuel stops told me that the nuclear heat source was as close to perfect as I could conceive with available technology. My main concern was what I had always been told was the reason that the Navy had limited its use of nuclear power - I thought that the fuel was prohibitively expensive.
In addition to the fuel availability and cost challenge that I thought existed, I knew that making use of that concentrated nuclear heat with the systems available in the 1950s and 1960s required high pressures, careful chemistry control, thick walled piping, phase changes, air tight cold water piping, and rather substantial fresh water storage tanks to make up for any plant losses. I had crawled the piping systems and signed enough work packages to know the plant almost by heart. I was pretty certain that I had a reasonable basis for knowing the source of much of the construction, design and operational cost.
I am not sure exactly when it happened or what information source opened my eyes, but I soon discovered that uranium based fuel, even when it is carefully engineered and manufactured, is far cheaper than the distillate oil burned in most combustion turbines. On the basis of heat content, commercial nuclear fuel cost less than 1/6 as much as distillate fuel - even in the early 1990s.
There were still sources that claimed that the highly enriched fuel used for naval propulsion systems was an exception to that rule, but then I found out that at least one marine propulsion reactor - the one that powered the NS Savannah - was fueled with low enriched uranium that was almost identical to commercial nuclear fuel. Even with that limitation, the Savannah's core was compact - especially compared to fuel tanks - and provided a reasonably long 5 year cycle between refueling.
Here is a quote from the Savannah's National Historic Landmark Nomination, a document that was not available to me at the time that I was learning about low enriched propulsion reactors. It provides similar information to the documents that I found buried in the USNA library:
Within the larger containment vessel, the reactor itself was housed within a "primary shield." This shield was a water-filled, 17' high, 2" to 4" thick lead tank. The reactor's active core was a circular right cylinder 62" in diameter and 66" high. The core was made up of 32 fuel elements. Each fuel element comprised 164 stainless steel fuel rods, .5" in diameter. The rods contained uranium oxide pellets, enriched to an average of 4.4 percent of U-235. The fuel rods in the centermost 16 fuel elements contained uranium oxide at an enrichment of 4.2 percent U-235, and in the outer 16 fuel elements the enrichment was 4.6 percent U-235. This compares to the longer lasting, 90 percent enriched uranium used in Navy reactors. Savannah's uranium oxide pellets, were .4255" in diameter, and the space between the pellets and the inner tube wall contained helium gas under pressure to assure good heat transfer across the fuel rod.I also discovered that the US had already shut down one of its enrichment plants (Oak Ridge) and was considering shutting down one of the two remaining facilities because there was too much enrichment capacity in the world market. That situation did not make it sound like enriched uranium was a scarce, inherently expensive resource based on my understanding of Economics 101.
Simple, light-weight, series-produced heat enginesAfter identifying the main cost drivers for the nuclear steam plants that had fallen out of favor with both the general public and the Navy surface fleet - with the exception of the aircraft carriers, I turned to evaluate Brayton Cycle gas turbines, the heat engine that was all the rage in the electric power industry. Not only were the merchant power producers and the electric utilities excited about machines that were durable, lightweight (compared to steam turbine systems), easy to build, cheap to install and simple to replace, but the Navy surface fleet had made the same decision for many of the same reasons.
(Source: Wikipedia Jet Engine Diagram under creative commons.)
Ship propulsion gas turbines look like the engines that hang under the wings of jet aircraft. That resemblance has a logical basis - technically speaking there is little difference. In fact, the LM-2500 workhorses that power frigates, cruisers, destroyers and now some amphibious ships are classified as aero-derivative (modified jet engines) gas turbines. Combustion gas turbines are simple machines that have just a major components - there is a compressor, a heat source (burners) and an expansion turbine that powers the compressor.
For jet engines, the energy left over after the compressor drive turbine exits at high velocity to produce thrust; in turbo props, generators and ship propulsion engines, hot gases for thrust are not particularly useful. Instead those systems need a slight modification in the form of an additional power turbine. Instead of shooting high energy gas out the back end of the engine to push a plane, ship engines and power generators use that high energy gas to spin a turbine connected to a propeller or a generator - through reduction gears, if necessary, to match the optimal rotational speed.
All heat engines need a heat sink; for steam plants the heat sink is a condenser cooled by water flow. For Brayton Cycle combustion gas turbines, the heat sink is the atmosphere that accepts the hot exhaust gases. After turning a power turbine or providing thrust, the hot exhaust gases are back to atmospheric pressure but they are far hotter - and more contaminated - than the air sucked in by the compressor. With the exception of the burners where fuel is injected at a regulated rate, I recognized that there were no heat exchangers in the vast majority of combustion turbines in commercial service.
Systematically, ejecting hot gases and ingesting colder air connects the bottom curve on an h-s diagram and matches the same function as the condenser in a Rankine Cycle steam plant. Since no one has to pay for the atmosphere, that heat sink costs the combustion turbine owners and operators a lot less than the multi-tube condensers and sea water systems used in steam plants. Of course, on board ships and in most power plants using gas turbines, the heat sink portion of the system is not free - intake and exhaust stacks require metal, construction and design work, and some amount of routine maintenance. On board ships, intake and exhaust stacks consume valuable midships space in multiple levels. For military use ships, the exhaust stack on a gas turbine ship is a detectable vulnerability due to the heat and contaminants contained in the gases.
As I thought about the basic characteristics of both uranium fuels and Brayton Cycle gas turbines, I kept getting disappointed by the papers and drawings that I was seeking and finding on the topic of high temperature gas reactors with closed cycle gas turbines. The idea is deceptively simple. Replace the burner portion of a combustion gas turbine with a reactor heat source and add piping and a cooler between the turbine exhaust and the compressor to close the cycle with a heat sink resembling a steam plant condenser.
Most of the paper authors writing about closed cycle gas turbines though it would be better to add "refinements" to the system. Their papers describe the ideas for adding components and changing cycle parameters to increase the thermal efficiency by adding stages of reheating and intercooling, reduce the cross-sectional area of the machinery by increasing system pressure, and flatten the efficiency versus power output curves by using system pressure changes to control power output. There was nearly unanimous agreement that helium, a lightweight, hard to contain gas was the best choice as the working fluid and heat transfer gas. The HTGR/CCGT advocates described designs to each other in ASME or ANS sessions that would allow construction of "commercial" plants with turbine power output in the several hundred to one thousand Megawatt electric range, even though the largest combustion gas turbines in commercial service were significantly smaller than 200 MWe.
Throughout the 45 year period I studied - from 1946-1990, there were less than a handful of actual machines using high temperature reactors with closed cycle gas turbines constructed, but there were hundreds of papers and dozens of special sessions held on the topic at technical gatherings like IAEA, ASME and ANS meetings. During that same period, nuclear plant vendors built hundreds of light water reactors while simple cycle combustion turbine manufacturers refined their systems through dozens of generations that resulted in an installed base of millions of machines in markets as diverse as corporate turboprops, high speed hovercraft, large military ships, land based power generation, and Army tank engines.
The first mystery worth solving was why had this situation happened? The second was to determine if there was a path forward that could alter the situation and provide a real opportunity to reach the ultimately simple, low capital cost machine that could run on high energy, low cost, abundant, pollution free fuel.
Keep it Simple, Stupid.I have learned over time to favor systems designed under the principle of KISS. From the point of view of a former operating engineer, atomic gas turbine researchers had long since departed from this advice.
The complicating component additions offered by closed cycle gas turbine advocates like C. Keller and D. Schmidt in their 1967 ASME paper titled Industrial Closed-Cycle Gas Turbines for Conventional and Nuclear Fuel illustrates how the small community of people interested in the topic gradually made choices that took them away from a development path that would allow them to reach their goals. It is important to understand that C. Keller presented his first paper on closed-cycle gas turbines at an ASME meeting in 1946, just a year after WWII ended, so this paper indicates his continuing interest in the subject for more than 20 years along with his professional interest in producing something new and publication worthy.
A later paper presented by R. Calvo and R. E. Thompson titled Compact Closed Cycle Brayton System for Marine Propulsion again illustrates how the closed cycle gas turbine researchers had gradually moved to a point where most accepted that any system that would be built would use pressurized helium gas, an intercooler between the compressor stages, a recuperator (heat exchanger) between the outlet of the compressor and the inlet of the reactor (heated by turbine exhaust gases), and a system of pressure reservoirs that would allow operators to add and or subtract helium gas from the system in order to control helium mass flow rate and thus system power output.
Each one of those refinements disturbed me. Heat exchangers were one of the banes of my existence, especially when operating in the warm and ecologically abundant waters in King's Bay, Georgia. High pressure gas systems are a big part of submarine design and operation, but we learned to respect the hazards and the cost of operating such carefully designed systems. There is a good reason that as much high pressure piping as possible is located in places that are not normally occupied by people. I had also learned first hand just how hard it was to reliably keep low atomic number, monatomic gases in their place - they find their way through the tiniest pores and gaps in sealing systems.
One of the things that enabled the success of the light water reactors connected to steam plants is that they followed the KISS principle at the time that they were designed. They did not reach for ultimate efficiency, they did not try using multiple technologies without a track record, and they always seemed to be designed with the operators and maintainers in mind. My quest was to determine if there were any inherent reasons that the combination of high temperature reactors and closed cycle gas turbines led people to make things so darned complicated.