Steam continues to be the most common medium for process heat transfer and power generation in the world. It is popular because of its ease of generation, usage, flexibility and predictability and, most of all, because it is still the most economic method of providing heat to a process.
However, the cost of steam has been rising steadily with increasing fuel costs and often forms a major part of the total cost of utilities ranging from 8 to 25% of the total cost of utilities.
Why meter steam?
To manage the cost of steam, the plant energy manager needs to know where steam is used, how much is used, whether it is being used effectively and how its consumption can be optimised. The major reasons for using steam meters include:
Plant efficiency
Most processes that use steam have norms for the ratio of steam usage to the amount of product produced. A steam meter can indicate how efficient a particular process is compared to the norm, whether idle machinery is switched off and whether steam usage practice is within acceptable limits.
Energy efficiency
A steam meter will help to determine the efficiency of generation of steam in a boiler and, therefore, to tune a boiler to deliver the highest amount of steam per unit of fuel used. Losses in distribution of steam through leakage or equipment failure can also be monitored and controlled, resulting in energy savings which far outweigh the cost of metering.
Process control
Steam meters can indicate that the correct quantity of steam is being supplied to a process at the correct temperature and pressure.
Costing
Typically, steam is generated at a central boiler plant that acts as the supplier to different sections within a plant. The cost of any individual product can only be accurately computed if the steam cost debited to it is based on the actual quantity of steam consumed in its production and not the total steam cost amortised over the various products.
Custody transfer
Often, a plant or process buys the steam from a nearby plant or utility provider - with the steam meter determining the bill.
Steam fundamentals
Nearly all steam used for industrial processing and heating is saturated steam because it has the benefit of higher heat transfer. Saturated steam is generally considered to be a two-phase fluid since it can, in its saturated state, exist both in gas (vapour) and liquid (water) form simultaneously.
Dry saturated steam follows a relationship between volume and pressure as well as temperature which, though not linear, is predictable. This relationship, defined by steam tables (Table 1), allows a user to determine the specific volume, pressure and temperature - even if only one of these quantities is initially known.
Superheated steam is used mostly in power generation since it is imperative to have completely dry, clean steam driving the turbine. Superheated steam can be treated almost as a true gas since the elevated temperatures at which it exists evaporates all moisture. Consequently, it does not follow the same relationship as saturated steam and its specific volume depends on both pressure and temperature (or degree of superheat).
In industry, the steam is normally homogenous wet steam with the 'wetness' or amount of water present defined by the 'Dryness fraction' - usually 0,95 or better. In cases where the dryness fraction is less than 0,95, there are likely to be errors in metering, possible damage of steam metering devices and, worst of all, water-hammer in the pipes.
Conditioning of this wet steam with the use of efficient steam trapping, moisture separators etc, can impove the dryness fraction.
A key factor in efficient steam distribution is the design of efficient pipe lines. The size of a steam pipe is selected after taking into account the velocity, pressure drop due to friction, etc. Oversized pipes mean expensive installations and low steam velocities - resulting in higher heat losses due to radiation and condensation and poor quality steam. Undersized pipes cause higher pressure drops, high pipeline velocities, erosion of pipeline components, vibration and noise and the possibility of water-hammer. Consequently, the velocity of saturated steam is generally specified between 15 and 40 m/s.
The basics of steam metering
Measuring the velocity of a fluid in a pipeline and multipying this by the cross-sectional area of the pipe results in a volumetric flowrate:
Qv = v.A
where:
Qv = volumetric flowrate (m3/s)
v = velocity (m/s)
a = cross-sectional area (m2)
And by integrating the output over a period of time, the volume passed in a specific time can be derived. However, the volumetric flowrate or total volume of steam is of limited use since steam is costed on a quantity basis (ie cost per 1000 kg) and, in fact, what is required is a mass quantity meter or mass flowrate reading.
Unless the pressure is held absolutely constant at the point of metering, the density of steam will vary with changes in pressure. It is, therefore, necessary to determine the steam density in order to apply 'density compensation' so that the meter indicates mass flowrate (Qm). In order to do this, the pressure and/or temperature of the steam must be determined. Since saturated steam follows a fixed relationship between density, temperature and pressure, either pressure or temperature measurement is enough to provide density compensation. However, in the case of superheated steam, both pressure and temperature need to be measured to calculate the density and apply the correction.
If density compensation is not introduced, inaccuracy in mass measurement will inevitably result. For example, a 2% variation in pressure (and corresponding density variation) may lead to a 1% mass flow error. Even in well controlled practical steam systems, pressure and temperature can vary by 10% or more, giving rise to errors far greater than typical accuracy specifications of flowmeters.
Types of steam meters
The flowmeters most commonly used for steam metering are: orifice plate; variable area meter; turbine meter and vortex shedder.
Orifice plates
Almost 80% of all meters installed on steam are orifice plates used in conjunction with differential pressure transmitters.
Although the orifice plate system is simple and rugged with reasonable accuracy, its turndown is limited to 4:1, it can buckle due to water-hammer and it is liable to silt up. Further, the square edge of the orifice can become blunt (particularly if steam is wet) and this will change the coefficient of discharge, thereby affecting system accuracy.
Variable area meters
This type of flowmeter is simple and robust and has a linear output which results in a high turndown. However, the tube must be mounted upright, it has only moderate accuracy and limited flowrates.
Turbine meters
The turbine meter is extremely accurate over a wide turndown but wear or fouling of surfaces will lead to a need for frequent recalibration. Rotor bearing life is also a factor in reliability and these type of meters are fairly expensive.
Vortex shedders
Approximately 16% of all meters used on steam are vortex meters and this is expected to increase significantly in the next five years.
The main advantages are that over a range of flows, the rate (or frequency) of vortex shedding is proportional to the flowrate and the output is a digital primary. Further, it offers reasonable turndown, very high accuracy and no moving parts.
Special requirements of steam meters
While flowmetering is a well established science, the measurement of steam flow places special demands on the devices and methods used. They need to be tolerant of the high velocities, pressures and temperatures, offer good stability over the long term and must be economically priced. As a result, the following parameters need careful examination:
Density compensation
In an ideal world, the pressure in process steam lines would remain absolutely constant. Unfortunately, this is rarely the case - with varying load demands, droop and process parameters all contributing to pressure variations which in some cases can be considerable.
The majority of steam metering systems currently used for monitoring process steam usage do not have in-built density compensation and are specified to operate at a fixed line pressure. Even small pressure variations can dramatically affect meter accuracy.
The error in mass flowrate calculation of a vortex meter is given by:
Thus, if a vortex meter without density compensation is specified to operate at 10 bar (specific volume = 0,198 m3/kg from the steam table) but is actually operating at 9 bar (specific volume = 0,219 m3/kg), the error is +10,6%.
Dryness fraction:
Although density compensation takes care of the inevitable fluctuations and variations in the steam pressure, this assumes that the steam is totally dry. However, where the steam is not totally dry (almost all cases!) then the density compensation will not be correct.
As the density will be higher than that of dry saturated steam, the indicated flowrate will be lower than the actual value. The relationship between dryness fraction and mass flowrate is expressed by the equation:
where:
Qx = mass flowrate at a dryness fraction x;
and
Qd = mass flowrate at dry saturated conditions.
This gives rise to the errors as shown in Figure 1. Consequently, compensation for pressure variations must have the facility to set a dryness fraction figure. Although it is difficult to establish the figure, conditioning the steam supply ahead of the meter using a suitable moisture separator will improve the steam quality to a nearly predictable value.
Turndown
Since varying loads can give rise to wide flow variations, it is essential the steam flowmeter is able to cope with these. It should also cope with the typical velocities that are encountered in steam systems.
Although the flowmeter may be sized to give wide turndowns, it is important to check whether velocity at the minimum and maximum flows remain within the limits of steam practice - 10 to 50 m/s. Often, vortex flowmeter manufacturers do not take this into account while sizing meters for steam.
Figure 2 shows a typical demand curve for a distributed steam system with a high start-up load and variable demand through the day. An orifice plate meter, with a 4:1 turndown is sized on the peak load of 1000 kg/h. Any flowrates below 250 kg/h are 'lost' or, at best, recorded with a significant error.
Calibration
The accuracy of any flowmeter is largely dependent on its calibration method. Most flowmeters are calibrated on mediums other than steam and a correction factor is applied to the calibration. Although in principle this is acceptable, it is important that the flowmeter be calibrated on a medium which has the same phase as that medium on which it is to be used and at similar velocities. In the case of steam, the meter should be calibrated on air which is a gas. Further, the master meter against which the flowmeter is calibrated must have an accuracy at least five times that of the rated accuracy specification of the meter under calibration.
| Tel: | +27 11 314 1391 |
| Email: | [email protected] |
| www: | www.za.krohne.com |
| Articles: | More information and articles about KROHNE |
© Technews Publishing (Pty) Ltd | All Rights Reserved
printer friendly version