Measurement elements

(a) General characteristics

Certain terms, relating specifically to measurement, are in common use and are described briefly in the following:

(ix) Time constant. This is the time taken for an exponential expression to change by a fraction 0.632 (= 1 — e’[3]) of the time needed to reach a steady-state value.

(x) Transducer. A device that translates changes in the value of one variable to corresponding changes in another variable.

(b) Temperature measurement

The commonest elements in use are as follows:

(i) Bi-metallic strips. A pair of dissimilar metals having different coefficients of linear thermal expansion are joined firmly together and hence bend on change in temperature, making or breaking contacts or moving a flapper over a compressed air nozzle. To increase mechanical movement bi-metallic strips may be wound in helical or spiral form. Accuracies are about 1 per cent and the range of suitability is -180°C to 540°C. This is not really a very accurate technique and is unsuitable for use at a distance.

(ii) Liquid expansion thermometer. A bulb filled with a liquid having a suitable coefficient of volumetric expansion (e. g. alcohol or toluene) is connected by a capillary to bellows or a Bourdon tube. The former produces a linear motion to make contacts or move a flapper, in response to thermal expansion, whereas the latter, comprising a spirally wound tube, unwinds on rise of temperature to give a rotational movement at its centre to operate an indicator or a controller. Accuracy is about 1 per cent and the range of application is -45°C to 650°C. Long capillary tubes through ambient temperatures other than that at the bulb introduce error that requires compensation. The practical limiting distance between measurement and indication or other use is about 80 m with the use of repeater transmitters or 30 m without. The fluid content of the bulb is critical and depends on the temperature range of the application: insufficient liquid may evaporate at higher temperatures to exert a saturation vapour pressure. In other cases, excessive hydraulic pressure may damage the controller if the instrument is used beyond the upper limit of its range.

(iii) Vapour pressure thermometer. A liquid-vapour interface exists in the bulb and the saturation vapour pressure transmitted along the capillary depends only on the temperature at the bulb. No capillary compensation is therefore needed. Accuracy is about 0.5 per cent and the range of suitability is from -35°C to 300°C. The volatile liquids used to fill the bulb are usually toluene, alcohol and methyl chloride. These thermometers are generally only for use either above or below a given ambient temperature: if this is crossed there is the risk of the vapour condensing within the capillary and significant error occurring. By using a combination of immiscible, volatile and non-volatile liquids in the fill it is possible to deal with this difficulty.

(iv) Gas-filled thermometer. If the fill is above its critical temperature over the whole range of temperatures in the application it is a gas and its volumetric thermal expansion follows Charles’ Law, since it operates at a virtually constant volume. Absolute pressure within the system is then directly proportional to absolute temperature. The range of use is -85°C to 540°C.

(v) Thermocouples. An electric current flows through a pair of dissimilar metals joined to form a loop if one junction is hotter than the other. The current generated is used as the feedback variable and, depending on the metals adopted, the application is from -260°C to 2600°C. In practical applications the reference (cold) junction is kept at a constant temperature by automatic compensation using a thermistor in a bridge circuit.

(vi) Resistance thermometer. The variation in electrical resistance of a conductor with temperature change is used to provide feedback over a range of applications from -265°C to 1000°C. Accuracy is about 0.25 per cent. A Wheatstone bridge circuit is commonly used, one resistance being the temperature sensing element. Remote set­point adjustment and other refinements are readily possible. This instrument is simple, accurate, stable and has a virtually linear response.

(vii) Thermistors. The measuring elements are oxides of metals with an inverse, exponential relation between temperature and electrical resistance. Their response is rapid and although it is non-linear it is possible to reduce the non-linearity by combining them with resistors. Application is limited to a maximum temperature of 200°C.

(viii) Infra-red thermometers. These respond either to thermal radiation or light falling upon a detector and usually give a visual display. For thermal radiation the detector may be a thermopile, composed of suitable metallic oxides, a thermocouple junction, or a semiconductor. The response is delayed according to the time lag of the thermal capacity of the detector. With light-responsive detectors a change in the electrical property of the detector is caused by the absorption of light. The field of thermal radiation caused by adjoining surfaces must be considered and the nature of the surface (emissivity) of the object emitting the radiation affects the readings.

(c) Humidity measurement

(i) Mechanical. Human hair, parchment, cotton, moisture-sensitive nylon or polymer and other hygroscopic materials vary in length as their moisture contents change according to fluctuations in the ambient relative humidity. The mechanical movement that occurs is used to provide the feedback. Response is mostly non-linear, except in the middle range of humidities. Other hygroscopic materials cannot match the performance of moisture-sensitive nylon or polymer. Range 5 per cent -100 per cent, ±5 per cent.

(ii) Electrical resistance. A double winding of gold or platinum wire over a cylinder of suitable material is coated with a hygroscopic film of lithium chloride. Changes in the relative humidity alter the electrical resistance of the lithium chloride between the wires and this is used to provide the feedback to the rest of the control system. Accuracy is about ±1.5 per cent but deteriorates with age. Elements of this type only cover a band of approximately 20 per cent change in relative humidity so multiple elements must be used for control over a wider range. Other versions employing graphite and other materials are available but with poorer accuracies, of the order of 2 to 3 per cent.

(iii) Electrical capacitive. A thin film of non-conductive polymer material is coated on both sides of metal electrodes which are mounted inside a perforated plastic capsule. The relationship between humidity and capacity is non-linear with rising humidity. The non-linearity is dealt with electronically and the sensor is compensated against temperature in an amplifier circuit, to give an output signal in the range 0 per cent to 100 per cent relative humidity with an accuracy of ±2 per cent. Many other humidity sensing techniques are available but capacitive sensors appear to be most popular.

(d) Dew point measurement

(i) Direct reading. A sample is passed through an observation chamber, one side of which is of polished metal. A thermocouple measures the temperature of this surface. A second chamber, on the other side of the polished metal, contains evaporating ether

Which cools the surface (under manual control). Dew is observed to form on the polished surface and the thermocouple then indicates the dew point temperature of the air being sampled and passed through the observation chamber.

(ii) Indirect type. A pair of gold wire electrodes is wound over a cloth, impregnated with lithium chloride and covering a tube or bobbin. The bulb of a filled system or the element of a resistance thermometer measures the temperature of the air within the bobbin, the dew point of which is to be determined. When a low voltage is applied across the electrodes current flows between them through the lithium chloride on the cloth, generating heat and evaporating water from the cloth. The concentration of lithium chloride increases and its electrical resistance alters. When the cloth is dry the salt is in a solid crystalline state and becomes non-conducting. Moisture is re­absorbed from the air in the bobbin because of the hygroscopic nature of lithium chloride, and a current flows again. A stable temperature within the bobbin is reached when the thermal and electrical fluxes are in balance. At this condition it is a property of lithium chloride that the air inside the bobbin is at about 11 per cent relative humidity. This is then related with the measured temperature to give an indication of the dew point. The range of suitability is from -45°C to 70°C with a minimum humidity of 12 per cent. The accuracy is about 2 per cent. Performance is critically dependent on the cleanliness of the air. The best accuracy is in the range -23°C to +34°C and above 40°C, dew point. Contamination spoils accuracy and the response is slow.

(iii) Calculation based on measurement of temperature and vapour pressure. A resistance element measures temperature and, simultaneously, a capacitive element in the form of a thin film of polymer absorbs water vapour molecules. The rate at which the molecules are absorbed is directly proportional to the partial pressure of the water vapour. Dew point is calculated electronically. The output is dry-bulb temperature, vapour pressure, dew point and moisture content. Accuracy of ±2 per cent is claimed in the range -60°C to + 80°C dew point. Stability is also claimed and the output can be related to a digital control system through a microprocessor.

(e) Enthalpy measurement

Measurements of dry-bulb temperature and relative humidity, made simultaneously, are

Interpreted as enthalpy. Enthalpy can also be calculated from measurements of temperature

And humidity, and displayed or used for control purposes.

(f) Flow rate measurement

Several methods have been used, principally as a safety cut-out to protect the compressors

Of water chillers if the water flow rate decreases.

(i) Orifice plates. A standard orifice to a BS specification with pressure tappings located in the correct places is fitted in the pipeline and the pressure drop provides a flow rate indicator. A reduced flow rate gives a fall in pressure drop which is used to effect the cut-out. The method is not satisfactory when used with chilled water because scale and dirt, always present in a system, collect at the orifice and gradually increase the pressure drop, implying that the flow rate is increasing when, in fact, the reverse is the case.

(ii) Spring-load blade. A flexible metal blade is immersed in the pipeline and set during commissioning to spring back and make an electrical contact that stops the compressor if the water flow rate diminishes. Difficulties in setting the blade during commissioning have arisen.

(iii) Vortex shedding flowmeter. For the case of turbulent flow the presence of a blunt obstruction in a pipeline causes the formation of vortices which grow larger and are shed as flow occurs past the obstruction. The frequency of vortex shedding is proportional to water velocity and hence to the flow rate, if the relevant cross­sectional area is known. The Reynolds number must exceed 3000. Problems with the viscosity of the water have occurred when used with water chillers.

(iv) Venturi meter. A properly made venturi meter, with the correct pressure tappings, is a preferred solution. No orifice or blade is present and the entry and exit sections of the venturi tube taper smoothly to the throat, with little opportunity for dirt or scale to collect. A reduction in the flow rate through the throat increases the static pressure therein. Obstructions upstream and downstream cause errors and the downstrem section should expand gradually to recover velocity pressure and reduce the overall loss (see section 15.4). The meter should be fitted in long straight lengths of piping. There are other more complicated methods of flow rate measurement.

(g) C02 measurement

This has become an accurate process and can be adopted as a method of indicating probable indoor air quality and controlling the rate of fresh air supplied. The most commonly used technique for this purpose is by Non-Dispersive Infra Red (NDIR) sensors. Carbon dioxide strongly absorbs radiation at a wavelength of 4200 nm when illuminated by an infra-red light. Air is sampled into a chamber formed by a pair of parallel, gas-permeable membranes and the chamber is flooded with infra-red radiation. An infra-red detector at the end of the chamber measures the amount of radiation absorbed at 4200 nm and transmits a signal to the control system. Accuracies from 150 to 50 ppm are claimed and the response time can be quite short, but it depends on the instrument used. Calibration may be held for months but dirt, condensation and mechanical shock can spoil this, according to ASHRAE (1997). Other techniques are also used.

Posted in Engineering Fifth Edition