The conversion of primary energy (e.g. heat generated by the combustion of a fuel) into useful work most often involves a working fluid operating in a thermodynamic cycle. Examples include a steam turbine operating in the Rankine power cycle to drive an electrical generator and a refrigerant circulating between the compressor, condenser, expansion device, and evaporator of a reverse-Rankine refrigeration cycle in your home or car air conditioner. Designing such equipment to operate safely, reliably, and at optimal efficiency requires accurate properties of the working fluid. Our goal is to provide accurate properties and models for working fluids.
We apply a wide range of experimental apparatus to the task of measuring the properties of working fluids. Many of these instruments are the same as those used in our work on standard reference fluids, and our data for working fluids approach the accuracy of the reference fluids. Thermodynamic cycles operate over a range of temperature and pressure and usually involve both liquid and vapor phases, and thus our measurements must cover similar broad ranges. The thermodynamic properties (density, vapor pressure, enthalpy, heat capacity, phase behavior, etc.) determine the operating conditions (temperature, pressure, flow rates, etc.) and efficiency of the cycle. The transport properties (thermal conductivity and viscosity) have a large effect on the heat transfer characteristics, which, in turn, greatly impact the cost and economic feasibility of equipment. The data we measure forms the basis of property models that are promulgated to industry through NIST Standard Reference Databases, such as the NIST REFPROP database.
Replacements for CFC and HCFC Refrigerants
Refrigeration and air-conditioning have become essential in our economy. For much of the 20th century the refrigerants (i.e. working fluids) used in the vast majority of refrigeration and air-conditioning equipment contained chlorine. R-12 (dichlorodifluoromethane or CFC-12) was ubiquitous in home refrigerators and automobile air conditioners. R-22 (chlorodifluoromethane or HCFC-22) was used in most small domestic and commercial air-conditioning systems, while R-11 (trichlorofluoromethane or CFC-11) was used in many large systems. In the 1970s it was first suggested that chlorine could, if transported to the stratosphere, destroy ozone that protected the earth's surface from harmful solar ultraviolet radiation. This hypothesis was proven in the 1980s, leading in 1987 to an international treaty known as the Montreal Protocol that regulated CFCs and HCFCs. An intense industry-wide effort to find and implement alternatives with zero ozone-depletion potential continued throughout the 1990s.
The Thermophysical Properties Division embarked on a major research program to supply the property data needed to select the best alternatives and design equipment equipment using them. Many fluids were investigated, including all of the now-familiar "new" refrigerants; several that seemed promising, enjoyed a brief flurry of interest, but have returned to obscurity; and several of interest for niche applications. The work proceeded in several rounds. Preliminary measurements provided the minimal data needed to fit a first equation of state. This was soon followed by preliminary measurements and correlations of the transport properties. As it became obvious that a fluid would see widespread commercial use, additional measurements were carried out to expand and improve the formulations. NIST published more than 200 works on the refrigerants between 1987 and 2001, and our data comprise a substantial fraction of all of the data published on these fluids. Most of the alternatives were hydrofluorocarbons (HFCs) or blends of HFCs; your home refrigerator and car air conditioner now operate with HFC-134a, for example. Our work in this area was largely complete by 2003, but currently there is a push to replace the HFCs (see Properties of New, Low-GWP Refrigerants).
Properties of New, Low-GWP Refrigerants
While the HFC refrigerants have zero ozone depletion potential (ODP) they are potent greenhouse gases. Their total contribution to antropogenic global warming is currently small, but there is concern that this will only increase over time, and regulations restricting the use of HFCs are under consideration in the U.S. and starting to take effect in Europe. The currently available refrigerants with low GWP values have drawbacks that limit their use in many applications, such as flammability (e.g., propane), operation at very high (CO2) or very low (H2O) pressures, or toxicity (e.g. ammonia). In response, a new class of refrigerants based on fluorinated analogues of propene (propylene) is under development by industry. We have measured the properties of the first two of this class of fluids, namely R1234yf (2,3,3,3-tetrafluoroprop-1-ene) and R1234ze(E) (trans-1,3,3,3-tetrafluoroprop-1-ene). We have measured the vapor pressure, p-r-T (pressure-density-temperature) properties, and thermal conductivity of both fluids. Our colleagues in the Theory and Modeling of Fluids Group have used these data to develop property models, and both fluids are now included in the NIST REFPROP database. Our work comprises the most comprehensive and accurate experimental data and property models available for these two fluids. Other, similar fluids are under development, and this is expected to be a continuing research area for us.
McLinden, M. O.; Thol, M.; Lemmon, E. W. Thermodynamic properties of trans-1,3,3,3-tetrafluoropropene [R1234ze(E)]: Measurements of density and vapor pressure and a comprehensive equation of state. International Refrigeration and Air Conditioning Conference at Purdue, W. Lafayette, IN, July 12-15, 2010; paper 2189.
Working Fluids for Organic Rankine Cycles
We are working with United Technologies Research Center (UTRC) to measure the properties of candidate working fluids and fluid mixtures for Organic Rankine Cycles. This technology is similar to that used in a conventional steam-based power cycle except that it uses an organic working fluid instead of water to allow operation at lower temperatures, including geothermal or solar heat sources. In the present work, novel mixtures are being investigated for use as working fluids, but only limited property data are available for most of the mixtures proposed. It makes little sense to make extensive property measurements on working fluids before they are known to be suited for use in Rankine cycles, but their evaluation requires property data. To resolve this conundrum we are taking a multi-phase approach with feedback between the cycle analysis, fluid optimization, and thermophysical property portions of the work. Based on our preliminary (and often estimated) property data, our UTRC colleagues run cycle simulations to evaluate candidates. For fluids passing this initial screening, we then measure properties (density, speed of sound, and viscosity) of the best candidate pure fluids and pressure-temperature-composition (p-T-x) properties of candidate mixtures to allow the development of improved property models, which are passed back to UTRC for a final, more detailed evaluation.
Innovative power cycles are being developed for production of electric power with higher efficiency and reduced environmental impact, including the possibility of CO2 sequestration. Design of these systems requires properties of the gas mixtures passing through the power turbines and other parts of the cycle. These mixtures have a large water content, along with other gases such as nitrogen and carbon dioxide; they are at high temperatures where thermodynamic data are scarce and very difficult to measure. In a project funded by the Department of Energy and joint between the Experimental Properties of Fluid Group and Theory and Modeling of Fluid Group, we use computational quantum mechanics to develop accurate surfaces describing the potential energy between the molecules, from which the thermodynamic properties can be calculated. We are performing density measurements on the key water-nitrogen and water-CO2 systems at experimentally accessible temperatures (up to 500 ˚C) to validate the theoretical results; the theory will then be used to calculate properties at higher temperatures.
We are also measuring the thermal conductivities of these mixtures at high temperatures; in this case the measurements are essential because no good theoretical approach is available for this property. The measurements use the transient hot-wire method, which has become the method of choice for high-accuracy thermal conductivity measurements. Because water's electrical conductivity is problematic for the traditional DC implementation of this technique, an AC version of the apparatus has been developed.