ASME began working on developing its steam tables, which list standardized thermodynamic properties for water in its vapor, liquid, and supercritical states, at a meeting in Cambridge, Mass., on June 23, 1921. It took decades of building consensus before the first ASME Steam Tables could be published. Today, steam tables are so ubiquitous that most engineers take them for granted.
Innovations are generally celebrated as the production of something altogether new, whether it’s the lightbulb, the television, or the smartphone. There was a market for a new device, and an inventor (or team of industrial engineers) comes up with something that fills the gap.
ASME began working on developing its steam tables, which list standardized thermodynamic properties for water in its vapor, liquid, and supercritical states, at a meeting in Cambridge, Mass., on June 23, 1921. Paradoxically, the problem that the original ASME project was trying to solve was not a lack of steam tables, but an excess. Several books were available, and they disagreed with each other, especially at higher temperatures and pressures that were becoming important as boiler and turbine technology advanced.
From that first meeting, it took decades of work and building consensus before the first ASME Steam Tables could be published. Today, steam tables are so ubiquitous that most engineers take them for granted, oblivious to the work they represent. This centennial affords an opportunity to reflect on the history of these efforts and look toward the future.
If there was already a large body of research into the properties of steam, why did ASME need to get involved? After all, by the 1920s, the industrial world had been in the so-called Steam Age for some time. Locomotives and ships were still generally powered by steam engines, and steam turbines had been in use to generate electricity for over 30 years. (Even today, much of the world’s electricity is generated via steam turbines.) The world ran on steam power.
However, understanding the properties of steam was vital for improving the design of these critically important machines. In spite of the revolutions in physics in the late 19th and early 20th centuries, this understanding could not be accomplished via theoretical calculations. Steam is not a “perfect gas” and therefore does not obey the ideal gas laws, which means that no simple formula can calculate its properties over a wide range of conditions. Instead, the properties of water and steam must be determined through measurements in combination with thermodynamic relationships. Once the properties are determined at discrete points, formulas are developed to interpolate values between known points or to extrapolate the data beyond the measured points.
Starting in the 1840s, many efforts were undertaken to measure the properties of steam and develop tables that described the relationship between temperature, pressure, volume, heat content, and other properties. In 1906, Richard Mollier in Germany published what can be considered the first modern steam property table, introducing the concept of the property enthalpy in the process. Others who published steam tables in the first part of the 20th century include Hugh Callendar in England and the American team of Harvey Davis and Lionel Marks. Unfortunately, these different tables did not agree and over some portions diverged quite a bit from one another. A chart accompanying a 1925 article published in Mechanical Engineering magazine compared the saturated vapor enthalpy as a function of pressure from several contemporary tables; no two curves were alike.
The problem was not simply a matter of accuracy, although that is important. Scientific experiments with differing results are common and will work themselves out over time. The larger problem arose because these steam tables were used by engineers in contracting, where performance guarantees for equipment such as boilers and turbines could change drastically depending on the steam tables used. Commerce needs a level playing field to ensure that, for example, differences in guarantees from different vendors result from real performance differences rather than being an artifact of the steam tables used.
Given ASME’s long-time involvement with the steam power industry, it made perfect sense for the Society to develop a standardized steam table. Davis, who published a set of steam tables with Marks, was involved with organizing the 1921 meeting that kicked off that effort; he later served as ASME president. Thanks to its position within industry and academia, ASME was able to solicit funds from industry to support this work and sponsored experimental research at Harvard and MIT to obtain more accurate steam properties; important data were also obtained at the National Bureau of Standards.
Meanwhile, engineers in other countries came to similar realizations, and the need for international standardization became evident. International Steam Table Conferences were organized in London in 1929, Berlin in 1930, and New York in 1934. The 1934 conference, held under the auspices of ASME’s Special Research Committee on the Thermal Properties of Steam, led to an agreement on a set of “skeleton tables” that could serve as a worldwide basis for water and steam properties. The landmark 1936 book, Thermodynamic Properties of Steam Including Data for the Liquid and Solid Phases by Joseph H. Keenan and Frederick G. Keyes, was consistent with those 1934 tables and served as the de facto standard steam tables in the U.S. for decades.
In the 1950s and 1960s, technological innovations required data for steam at pressures and temperatures beyond the boundaries of existing tables, while the increased use of computers in engineering called for steam properties to be related in the form of equations rather than by interpolation from printed tables. Efforts of an International Formulation Committee resulted in the adoption in 1967 of a standard called IFC-67, which divided pressure-temperature space into six regions with different equations. IFC-67 also served as the basis of the first official ASME Steam Tables, some 46 years after the ASME effort started.
Keeping It Current
Shortly after IFC-67 was adopted, stakeholders decided that a permanent organization was necessary to manage international standards for thermophysical properties of water and steam. The International Association for the Properties of Steam (IAPS) was officially established in 1971; in 1989 the name was changed to the International Association for the Properties of Water and Steam (IAPWS) to reflect increased interest in water and aqueous solutions for applications beyond the steam power industry. IAPWS is structured as a collection of national committees; the U.S. representation is provided through the Properties Subcommittee of the ASME Research and Technology Committee on Water and Steam in Thermal Systems. This committee, which traces its history to the original ASME steam tables meeting in 1921, consists of representatives from the power industry as well as researchers from academia and government.
The current “steam tables” exist on two separate tracks. For many purposes, a reference formulation with the highest possible accuracy is needed, and computational speed is not important. In 1995, IAPWS adopted a “formulation for general and scientific use,” which has informally become known as IAPWS-95. At the time of its development, IAPWS-95 represented the state of knowledge over a wide range of conditions, up to 1,273 K and 1,000 MPa, and it is still the definitive reference for scientific work.
However, the needs of the power industry are different. IAPWS-95 is too slow for many iterative design calculations. Also, because changing steam tables affects many commercial calculations (such as contracted heat rates), industry needs a standard that is stable over a period of decades to avoid the trouble and expense of changing. This need is met by a separate “industrial formulation” adopted in 1997, known as IAPWS-IF97.
IAPWS-IF97 was fitted to the reference IAPWS-95 formulation to within small tolerances, with pressure-temperature space broken into five regions and special attention paid to computational speed and consistency at region boundaries. In essence, IAPWS-IF97 trades a small (often negligible) amount of accuracy and consistency for a speedup in calculations of two or three orders of magnitude. Additional speedup can be obtained with “backward” equations that allow calculations with what would often be a dependent variable as input, such as pressure as a function of temperature and entropy. Current steam tables books are based on IAPWS-IF97.
Additional speedup in restricted domains, such as for computational fluid dynamics (CFD), has also been studied by IAPWS.
Scientists and engineers need more than just the thermodynamic equation of state, so IAPWS has developed standard formulations for other thermophysical properties of water and steam, such as viscosity, thermal conductivity, surface tension, and static dielectric constant. In the case of the viscosity and thermal conductivity, simplified “industrial” calculation forms are available. Formulations have also been developed for the thermodynamic and transport properties of heavy water, which is of interest both in scientific applications and in certain nuclear reactor designs, and for the thermodynamic properties of ice.
In addition, IAPWS studies and issues technical guidance for such issues as the chemistry involved in preventing corrosion, deposition, and other related problems in power plants; makes recommendations for the Henry’s constants and vapor-liquid partitioning coefficients for common gases in water over a wide temperature range; and developed the first fully consistent thermodynamic equation of state for seawater— an advance for physical oceanography, which requires accurate and consistent representation of the properties as a function of salinity.
Over the Horizon
Some may contend that steam tables will become less important in the future, as more electricity is generated by wind farms, solar panels, and gas turbines, and there is less need for steam power equipment. Although the progress of alternative energy has been rapid, it is likely that steam turbines and combined-cycle power plants will exist in significant numbers in the coming decades. Even in a carbon-neutral future, water and steam properties will be needed for nuclear plants and in niches in the renewable energy domain. These applications will continue to require knowledge of properties such as water/steam distribution ratios and solubilities of solutes, along with the guidance on power cycle chemistry supplied by IAPWS Technical Guidance Documents.
In addition, the properties of water and steam will continue to be needed for science and industry outside of power generation, ranging from food processing and desalination to oceanography and planetary science.
Because of the roots of IAPWS in the power industry, it could be a logical place to develop standards for properties of alternative power-cycle working fluids such as carbon dioxide if such alternative cycles become widespread. The thermophysical properties expertise and institutional knowledge within IAPWS and ASME will be applicable for any setting where technical requirements dictate the necessity of accurate consensus standards for the properties of a fluid.
The current scientific standard for thermodynamics of water and steam, IAPWS-95, is likely to be replaced within the next 10 years to reflect new experimental data and better understanding of theoretical constraints.
Since IAPWS-95 is already quite accurate at the temperatures and pressures encountered in most applications, improvement will only be significant at low temperatures (metastable supercooled water), at high temperatures and/ or pressures, and perhaps near the vapor-liquid critical point. Replacement of IAPWS-IF97 may follow, but that is a more serious undertaking due to its widespread use for performance guarantees in the power industry.
IAPWS will continue to study other properties of water and aqueous systems. Examples of current efforts include the development of formulations for the viscosity of seawater and for the diffusivity of water and steam.
One growth area is molecular modeling. Ab initio quantum chemistry and molecular simulation show promise for calculating properties at conditions where accurate measurement is difficult. While we are still a long way from being able to calculate most water properties with better accuracy than they can be measured, we are nearing the point where molecular calculations can provide important supplementary data to improve correlations, especially in guiding extrapolation where good data are lacking.
It might be tempting to think that advances in computing power would render superfluous work to calculate accurate properties more quickly, but we do not agree. Engineers will always want to push the limits, whether that means enabling more complex system design and optimization calculations or allowing a finer grid to be used to model steam flow through turbines using CFD. For the foreseeable future, anything that decreases the computational effort for steam properties by a factor of two, or a factor of 10, will continue to be welcomed in engineering.