1. Introduction

When making a decision regarding the selection of a pump from among several offers, one should not only focus on the lowest price criterion, but should also take into account the operating costs over several years. For this purpose, you can use the so-called LCC method [1], which has been gaining popularity in recent years. The main component of operating costs is the cost of electricity used to drive the pump. To illustrate this obvious point, let us give a computational example. Let us consider a pump with a capacity of Q = 2000 m³/h lifting height H = 50m. (These are parameters corresponding to double-stream pumps used in larger waterworks and heating systems). When pumping water, the so-called the effective power consumption of the pump with the given parameters (i.e. the power that a pump with 100% efficiency would consume) is 272.5 kW. Actual power consumption depends on its efficiency. For pumps with similar parameters designed and manufactured by reputable companies, the efficiency range that can be expected is between 83 and 88%. For a pump with an efficiency of 83%, the power consumption will be approximately 328 kW, and for a pump with an efficiency of 88% it will be 309 kW. The difference is therefore 19 kW, which in the case of pumping for 6000 hours a year and the energy purchase price of PLN 45/kWh translates into a PLN 51 difference in the costs of energy used. This represents a significant percentage of the price of a pump with such parameters, and therefore energy consumption should be taken into account as an important selection criterion. It should be emphasized again that in the above example, only pumps with a decent technical level were taken into account, and on the market there are also pumps with efficiencies that are less than the expected level.

This method of selecting a pump that takes into account the criterion of energy consumption, which is correct in principle, encounters problems in practice, as described below.

2. Efficiency of a specific pump

You should be aware that the efficiencies of individual pumps of the same type may differ significantly. This is due to the fact that the parameters, including the efficiency of the centrifugal pump, depend mainly on the geometry and roughness of the walls of the flow channels of rotors and vanes, and in the vast majority of pumps these are made using foundry technology. Typical foundry technology based on sand-based casting molds does not ensure full repeatability of geometry and surface roughness. This is due, among other things, to the fact that the liquid metal solidifying in the mold undergoes shrinkage, which is a largely random process, as are the resulting shape deviations. Standards regarding the tolerance of shapes and dimensions of castings sanction this state of affairs by allowing significant deviations. In order to improve the accuracy of castings, and thus to obtain better repeatability of pump parameters, more complex casting technologies can be used (e.g. ceramic cores, pressure casting, etc.) or the rotors can be made by machining, which is possible for open rotors but very difficult. for closed rotors. However, the use of this type of technology is associated with a significant increase in manufacturing costs and for this reason it occurs mainly for pumps of basic importance, while traditional foundry technologies dominate for general purpose pumps.

As an illustration, on Fig. 1 shown are the results of actual measurements of the characteristics of many copies of the same type of pump against the background of the tolerance field allowed by EN ISO 9906 standard and compared to the characteristics according to the Operation and Maintenance Manual.

Fig.1 Collective characteristics of many pumps of the same type.

It should be emphasized that the dispersion of the actual parameters of individual pumps does not result from the carelessness of their manufacturers, but to a large extent from the natural inaccuracy of commonly used technologies. This fact was taken into account in EN ISO 9906 standard, which defines the permissible tolerances of pump parameters. This standard defines two accuracy classes. The second class, considered standard if the ordering party and the supplier do not specify otherwise in the contract, allows pump head deviations in the range of +/- 5%, efficiency deviations in the range of +/- 8% and efficiency deviations in the range of -5%. In the first accuracy class, the tolerances are narrower and amount to +/- 3% for lifting height, +/- 4.5% for efficiency and -3% for efficiency.

Fig. 2 Criteria for accepting pump parameters according to the EN ISO 9906 standard.

The method of assessing pump parameters according to EN ISO 9906 is shown as a reminder fig. 2. Guaranteed pump parameters (capacity Q and lifting height H) define the warranty point. From this point, a cross is drawn in the drawing, the length of the arms of which corresponds to the Q and H tolerances given above. The pump is considered to meet the guaranteed parameters if its actual characteristics, measured in the manner specified in the standard, meet at least one of the arms of the cross. In the case of efficiency, a line is drawn from the origin of the coordinate system to the guaranteed point. The intersection of this line with the measured, actual pump characteristic determines the capacity for which the efficiency is assessed. So on Fig. 2 the measured efficiency at the location shown by the vertically downward arrow would be compared with the guaranteed efficiency.

As you can see, according to the standard, the actual pump efficiency may differ significantly from the guaranteed (offer) efficiency. Not only can it be 5% lower than the offer efficiency, but it can also be assessed at a completely different efficiency than the guaranteed one.

This results in the above-mentioned practical difficulties when selecting a pump based on the efficiencies provided in the offer. In an extreme case, a situation may arise where, for example, a choice is made between pump A and pump B, and the efficiency of the former is 4% higher, which, based on the calculation of the cost of energy consumption, results in a decision to choose pump A. However, it may prove that a particular example of pump B has an efficiency that is 1% higher than a particular example of pump A, and this is within the tolerances allowed by the standard.

It can be argued that despite the dispersion in the efficiency of specific units, there is a greater statistical chance of obtaining higher efficiency for the pump unit whose offered efficiency is higher. This would be the case if all companies used the same methodology for determining offer efficiency, but this is not the case because there is no standardized practice in this area. There is no doubt that the efficiency given in offers should be based on measurements of a number of pumps whose efficiencies, as mentioned, vary. However, there are various possibilities for determining bid efficiency based on measurements of a number of pumps. They are shown for illustration purposes only Fig. 3. Thin lines on Fig. 3 mean exemplary, measured efficiency characteristics of several pumps showing scatter resulting from the manufacturing technology (the characteristics of unsuccessful units with too low efficiency were rejected). The most reliable way to determine the efficiency declared in the offers would be to adopt a certain average efficiency, such as the one shown in Fig. 3 with a dashed line. In such a case, the specific unit delivered to the customer could have efficiency both higher and lower than the offer one. However, the maximum efficiency obtained on a specific unit, such as the one shown with a solid line, can be assumed as the offer efficiency. In this situation, the specific delivered unit would have an efficiency equal to or lower than the offer one, but within the tolerance range. It cannot also be ruled out that a company that has technology that ensures good repeatability of pumps (e.g. efficiency dispersion within 2%) will accept as the offer efficiency an efficiency that has never been obtained in tests, but is higher enough (e.g. 3%) that the efficiency of specific copies are below it but still within tolerance. This situation is shown in Fig. 3 dotted line. In such a case, the buyer will always receive a copy with a lower efficiency than the offer one. All described possibilities for determining the offer efficiency are not inconsistent with EN ISO 9906 standard, while the method adopted in a given company depends on the degree of aggressiveness of its marketing policy. It should be emphasized that all these methods, although differing in the degree of reliability towards the recipient, still enable the delivery of pumps within the tolerance resulting from the standard. In a publication published in a serious magazine, it is inappropriate to write that there may be companies misleading customers by providing significantly overestimated efficiencies, even outside the tolerance range. However, such practices may occur because it is not always possible to perform a measurement verifying the efficiency of the supplied pump in the installation, among other things, due to the difficulty of installing the flow meter. In his practice, the author encountered a case when a pump delivered by a reputable company had an efficiency that was 10% lower than the one offered.

Fig. 3 Methods for determining bid efficiency.

3. Recommended course of action

The disparity in the efficiencies of specific pumps and the lack of generally recognized standards for determining the offer efficiency means that the generally correct method of selecting a pump, taking into account the energy consumption criterion, may not work in practice, as the difference between the efficiency of a specific unit and the offer efficiency may eliminate the effects calculated on the basis of the offers. This does not mean that you should give up taking energy consumption into account when selecting pumps, but you should approach the efficiency quoted in the offers critically. Data exist to estimate the pump efficiency that can be achieved [2] depending on its efficiency and speed characteristics. If the offer includes a pump with an efficiency significantly different from the others and from the data from the literature, it is advisable to approach such an offer carefully. In such a case, it is advisable to include in the contract the testing of pump parameters by an independent entity and to impose appropriate contractual penalties if the performance of the offer is not met. An independent acceptance test of pump parameters is advisable whenever higher powers are involved.

The ordering party may request that the pump be accepted according to accuracy class I according to EN ISO 9906, which limits the spread of parameters. However, in such a case it should be taken into account that the prices of the offered pumps will increase. Contractors wishing to limit the spread of parameters must either anticipate the use of more expensive technologies (e.g. purchase of castings in a specialized foundry) or take into account the cost of correcting the parameters of a pump made using standard technology. Such correction, performed in the event of excessive parameter deviations found during tests, involves disassembling the pump and corrective measures (often performed manually), such as further cleaning or grinding of the flow channels, sharpening the edges of the inlet blades, correcting the dimensions of the sealing gaps, etc.

According to the author, ordering parties should request in their specifications not to include the efficiency with the tolerance resulting from EN ISO 9906 (in accuracy class I or II), but should require the determination of the minimum guaranteed efficiency (i.e. allow only positive efficiency tolerances), require acceptance tests (observed or conducted by an independent entity) and impose failure to meet the guaranteed efficiency with contractual penalties comparable to the losses resulting from increase in energy consumption costs. This approach would ensure that the energy savings estimated at the stage of comparing offers would actually be achieved, and the selection of an offer taking into account energy consumption costs would be based on comparable data. This would also result in a unification of manufacturers' approach to the method of determining offer efficiency, which in the event of a request to provide the minimum guaranteed efficiency in the offer should consist in adopting the minimum efficiency found in tests (after possibly rejecting clearly unsuccessful units).

4. Summary and conclusions

Choosing a pump that is generally correct, taking into account the power consumption criterion, may in practice not lead to the expected results due to differences in the efficiency of the delivered unit in relation to the efficiency offered.
It is advisable to require that offers include not the efficiency with tolerances resulting from the EN ISO 9906 standard, but the minimum guaranteed efficiency, as well as carrying out acceptance measurements of pump parameters and imposing appropriate contractual penalties for failure to meet the minimum guaranteed efficiency. The use of minimum guaranteed efficiency significantly increases the comparability of offers and the accuracy of energy consumption cost calculations carried out at the stage of evaluating and comparing offers.

Dr. Eng. Grzegorz Pakula

Literature

1. Strączyński M., Pakuła G., Urbański P., Solecki J. Manual of Pump Operation in Waterworks and Sewerage. Polish Waterworks Chamber of Commerce, "Seidel-Przywecki" Publishing House, first edition, Warsaw 2007.

2. W. Jędral, Centrifugal pumps, PWN Scientific Publishing House, Warsaw 2001.

The pump, like any machine, requires periodic renovations to improve its technical condition, which deteriorates during operation. Water supply pumps operate in relatively easy working conditions, which is due to the fact that clean, cold water is a relatively "friendly" medium, neither abrasive nor chemically aggressive. Sewage pumps have slightly worse operating conditions, but in this case the conditions are not extremely difficult. For this reason, the intervals between renovations for pumps in the water and sewage industry are long, especially for water supply pumps, where they can reach several or even a dozen or so years, provided that the operation is carried out correctly.

The decision to send the pump for overhaul is generally made for one of two reasons: either there are some mechanical problems (such as increased vibrations, noise, rotational resistance, problems with bearings or seals) or hydraulic problems, i.e. loss of the required capacity or lifting height. When making such a decision, economic criteria should also be taken into account, as it is possible that a mechanically efficient pump that still provides sufficient parameters already operates with reduced energy efficiency, which results in an increase in energy costs. Raising the efficiency to a level close to the initial level usually requires a relatively inexpensive medium overhaul, including replacement of sealing rings and regeneration of the rotor. It may therefore be profitable to perform this type of renovation relatively often, solely to maintain high efficiency.

One of the fundamental issues determining the level of pump operating costs is determining the optimal length of the period between renovations. Renovations carried out too infrequently increase energy costs due to reduced efficiency. On the other hand, carrying out renovations too frequently increases renovation costs excessively.

The interval between renovations can be determined by one of the following methods:

1. Work until failure occurs.
The pump is sent for overhaul only when it fails and prevents further operation. Contrary to appearances, this may be a rational method in some cases. For example, for small and cheap pumps with a low-responsibility role, especially when standby pumps are available, the introduction of monitoring and technical condition analysis systems may be disproportionately expensive in relation to the price of the pump. The advantage of this strategy is that renovations are certainly not carried out too early. However, the disadvantage is that if a failure occurs, the cost of renovation may be significant. A variation of this strategy is that for cheap, mass-produced pumps, repair is not profitable at all and after a failure, the pump is replaced.

2. Send for overhaul in accordance with the manufacturer's recommendations
Some manufacturers specify in the operating instructions after how long the pump should undergo a current, medium or major overhaul. The strategy of following these recommendations has the advantage of being clear and does not require expenditure on monitoring systems. This is also a safe strategy because manufacturers usually define the intervals between overhauls conservatively, so the risk of a failure occurring between overhauls, the repair of which would require significant costs, is limited. The caution of such an approach is also its disadvantage, because in practice it may lead to an increase in the costs of renovations due to the fact that they are carried out more often than would result from the actual technical condition of the pump.

3. Directing for renovation based on monitoring the technical condition
This method is based on monitoring selected pump operating parameters, such as bearing temperature, vibrations during operation or noise level. The pump is sent for overhaul when the parameters exceed the level considered critical. This is to carry out the renovation in advance, before it causes a failure that would be expensive to repair. Therefore, just like the previous one, this is a safe strategy, but more economical because it allows for a more precise capture of the moment when renovation is actually recommended. The disadvantage is the need to incur expenditure on the monitoring system, as well as the difficulty in formulating critical parameter values, for example determining what vibration level indicates the proximity of a failure. (We are not talking here about obvious situations, such as when the pump does not achieve the required efficiency, which indisputably qualifies it for renovation).
This method is particularly recommended for pumps with a complex structure, the repair of which after a failure is very expensive, and those that play a responsible role in the system, for example those whose failure to operate in an emergency may deprive the population of water supply.

4. Directing for renovation on the basis of economic calculation
This is a further development of the previous strategy in that not only the technical condition of the pump is monitored, but also the cost of pumping a cubic meter. It may happen that the pump, which is still operational, consumes an increased amount of energy due to deteriorated efficiency, which results in an increase in pumping costs.
To determine the optimal moment when the pump should be repaired, the number of cubic meters pumped and the amount of energy for pumping should be recorded. For analysis purposes, we assume that the cost of pumping a cubic meter consists of the renovation cost per pumped cubic meter and the cost of energy per pumped cubic meter. Other costs, such as the cost of ongoing maintenance, can be omitted because they are not related to the moment when the renovation takes place.

Let's assume that the cost of renovation to restore the original pump is known. If we divide this cost by the number of cubic meters pumped since the last renovation, we get the renovation cost per cubic meter pumped. Since we divide the constant value by the increasing number of cubic meters, this cost component hyperbolically tends to zero.

If we record both the amount of energy consumed by the pump and the amount of cubic meters pumped since the last renovation, then dividing the first value by the second we will obtain the average energy cost for pumping a cubic meter in this period. As the pump's efficiency deteriorates during operation, the cost gradually increases. For the sake of clarity, it should be noted that this reasoning is correct if the pump operates in similar conditions all the time. If its parameters and operating point change over time, the energy consumption per cubic meter changes more as a result of this than due to the deterioration of efficiency. The reasoning is also correct if the operating conditions change periodically, for example during the day, and the average parameters remain constant, because then the long-term energy consumption rate per meter is related to the deterioration of efficiency.

If we now add up both components of the cost of pumping a cubic meter, i.e. the renovation cost and the energy cost, at some point we will observe its minimum, because after a significant number of hours the renovation cost will become insignificant and the energy cost will increase. When we observe an increase in the total cost, a decision should be made to send the pump for renovation, because continuing to keep the pump in operation will result in losses due to increased energy consumption outweighing savings due to postponing the renovation.

This reasoning is correct under two conditions:
a) that the renovation cost is as expected
b) that as a result of the renovation, the pump will regain the assumed efficiency.

If the second condition is not met, i.e. the renovation does not increase the efficiency of the pump to the required extent, the renovation expenditure will be wasted. This indicates the above-mentioned need to commission the renovation to companies that will not only ensure that the pump achieves appropriate operational efficiency, but also, and even above all, will enable the restoration of hydraulic parameters, especially energy efficiency. After renovation, the pump should undergo tests including measurement of energy characteristics, which verifies the quality of the renovation. This strategy is especially recommended for pumps with significant energy consumption.

Dr. Eng. Grzegorz Pakula

 In practice, there is often a case where the pumping station's capacity must be regulated within a very wide range while maintaining an approximately constant pressure in the discharge manifold. A typical example is water supply pumping stations.

For the sake of clarity, it should be noted that the assumption of constant pressure in the collector is a certain simplification, because in practice, a change in efficiency causes a change in flow resistance in the pipelines in accordance with their characteristics, as a result of which an increase in pressure is required to obtain increased efficiency. For this reason, the pressure in the pumping station discharge manifold has a natural tendency to vary with capacity. To keep it constant, special pressure stabilization systems would be required. For the purposes of this article, for simplicity, we will assume the above-mentioned assumption of constant pressure in the collector.

If the efficiency is to vary within a very wide range, i.e. from the maximum value to almost zero, it is advisable to use a number of pumps operating in parallel because there is no method that allows for effective regulation of a centrifugal pump in the efficiency range of 0-100% while maintaining a constant discharge pressure. . The use of several pumps operating in parallel instead of one larger one has the disadvantage of reducing efficiency, because efficiency usually increases with the pump efficiency. However, it allows you to easily change the capacity of the entire pumping station by turning on and off the appropriate number of pumps. This type of regulation is of a step nature, which means that it allows for coarse regulation that allows obtaining capacities that are multiples of the capacity of one pump. Therefore, it requires "adjustment" in intermediate ranges. A common solution involves adjusting one of the pumps by changing the rotational speed. This pump is expected to add efficiency varying in the range of 0-100% to the other pumps (operating at constant speed and constant efficiency). This solution is used due to the low investment cost, as only one inverter is required, and the prices of these devices constitute a significant part of the investment costs. However, although this solution seems natural and logical at first glance, it is not technically correct, as will be shown in the example below.

First of all, we will remind you how to obtain the pump characteristics at variable speed, i.e. the so-called "shell characteristics". Of course, this problem does not occur if the manufacturer provides such characteristics obtained from measurements, which, however, is not always the case. In many cases, manufacturers only publish characteristics for a constant, nominal rotational speed. In such a case, if you want to predict the pump's operation at variable rotational speed, you can make a theoretical calculation of the characteristic. According to the similarity theory, when the rotational speed is reduced, the efficiency changes linearly and the lifting height changes with its square.

Therefore, if from the pump characteristic at rotational speed n₁ we will take any point with efficiency Q₁ and lifting height H₁ this when changing the rotational speed to n₂ we will obtain respectively:

                        Q₂ = Q₁ n₂ / n₁,

and: H₂ = H₁ (n₂ / n₁)².

It can be assumed that the efficiency at point (Q₂, H₂) for N₂ will be approximately the same as at the starting point (Q₁, H₁) for N₁. This assumption is justified for moderate changes in rotational speed, but when the speed is reduced too deeply, the efficiency deteriorates.

In this way, taking a few points from the characteristic for rotational speed n₁, you can obtain characteristics for other rotational speeds. Such a calculation is not 100% accurate because the similarity theory ignores many factors (such as the influence of wall roughness, etc.), but it allows to obtain predicted characteristics with an accuracy sufficient for practical analyses. (as mentioned, characteristics obtained from measurements would be more reliable and would not be burdened with errors resulting from the assumptions of the similarity theory, but they are not always available).

For example in Fig. 1 The pump characteristic is shown, which has a nominal point of Q = 1500 m for 400 rpm³/h, H = 50 m. This characteristic, in accordance with the formulas given above, was converted to 1400 and 1300 rpm. The change in the position of the nominal point is shown by arrows from dotted lines. As we know, this point moves along a parabola. The highest pump efficiency, without changing the value, shifts towards lower efficiency, proportionally to the reduction of the rotational speed.

Let us now consider the cooperation of two pumps of this type, one of which operates at a constant rotational speed of 1500 rpm, and the other operates at a variable rotational speed. (For simplicity, we ignore motor slip and assume that the characteristics of both pumps are identical). This case does not differ qualitatively from the case when instead of one there is a larger number of pumps operating at a constant speed.

As we know, the total (equivalent) characteristics of pumps operating in parallel are obtained by adding their capacities at a given lifting height. On Fig. 2 shown are the equivalent characteristics of two pumps operating in parallel at a speed of 1500 rpm, the equivalent characteristics of pumps operating in parallel at speeds of 1400 and 1500 rpm and pumps operating in parallel at speeds of 1300 and 1500 rpm.

Fig. 1. Pump characteristics at different rotational speeds.

Let's assume that the aim of the regulation is to change the capacity from 0 to 820 m3/h while maintaining a constant discharge pressure corresponding to 49 m of lifting height (i.e. approx. 4.9 bar). This value is marked on Fig. 2 horizontal dash line. The idea of ​​regulation in this case is that in the capacity range 0 to 410 m3/h one pump is turned off and the other covers this range at different rotational speeds. However, in the range of 410 to 820 m3/h, one of the pumps operates at a constant rotational speed, delivering 410 m3/h, while the other "supplements" the efficiency with the required value at variable rotational speed. In the case of constant discharge pressure (horizontal system characteristics), the regulated pump works identically both independently and in cooperation with the other. (It should be emphasized once again that in a case closer to reality, when the characteristics of the system due to the change in flow resistance are parabolic, the analysis would be slightly more complex). Therefore, it is enough to consider the operation of pumps in the range of 410-820 m3/h.

The maximum efficiency of 820 m3/h is achieved when both pumps operate at a speed of 1500 rpm. On Fig. 2 this corresponds to point A, where the equivalent characteristic of the two pumps intersects the horizontal characteristic of the system. Each pump separately operates at point D, which results from the intersection of the system characteristics with the individual pump characteristics for 1500 rpm. Point D is slightly beyond the nominal point but still within a favorable range in terms of efficiency.

Fig. 2. Parallel cooperation of a pump with adjustable speed and a pump with constant speed.

If we want to reduce the efficiency to 700 m3/h, we lower the rotational speed of the regulated pump to 1400 rpm. The operating point then becomes point B, located at the intersection of the system characteristics with the equivalent characteristics of pumps operating in parallel at speeds of 1400 and 1500 rpm. The constant speed pump continues to operate at point D. However, the operating point of the regulated pump moves to point E, which lies at the intersection of the system characteristics with the pump characteristic for 1400 rpm (shown in Fig. 2 thin line). Point E is in the performance range below optimal but still within the permissible range.

However, let us note what happens when the rotational speed of the regulated pump is lowered to 1300 rpm, which is required to reduce the total capacity below 500 m3/h. The two-pump system then operates at point C, which is the intersection of the system characteristics with the equivalent characteristics of pumps operating in parallel at speeds of 1300 and 1500 rpm. The constant speed pump continues to operate at point D. However, the operating point of the regulated pump moves to point F, which lies at the intersection of the system characteristic and the pump characteristic for 1300 rpm. This is the point at the beginning of the characteristic, in the performance range beyond the permissible range. The efficiency of the pump at this point, as seen from Fig. 2, is low, which means that the regulation is not optimal in terms of energy. This unfavorable effect could still be acceptable due to the fact that the pump, operating at low efficiency, consumes little power and therefore losses are limited. However, there are also unfavorable movement effects. When the pump operates at too low a capacity, it operates with an increased level of vibration caused by recirculation flows and with an increased radial force acting through the impeller on the shaft end. These effects reduce the life of the pump, including the life of its bearings. In his practice, the author has encountered cases of shaft breakage under the rotor hub, which are difficult to explain because they concerned pumps with a proven design that have been operating for many years without this type of failure. Such failures occurred on pumps operating as regulated ones, in parallel with constant speed pumps. This requires more careful demonstration, but intuitively one can suspect that the reason for the shaft failure was increased load due to operation in the prohibited area of ​​too low efficiency.

Shown on Fig. 2 the example concerns pumps with specific characteristics, but it shows the general effect that reducing the rotational speed of the pump, which is required to operate at constant pressure, causes it to "choke" and enter the area of ​​prohibited operation, which results in both operation with significantly reduced efficiency and and working in unfavorable traffic conditions.

How can this situation be remedied? If it is necessary to obtain a total efficiency slightly higher than the efficiency of a single pump, instead of using a parallel connection of a constant-speed pump with a reduced-speed pump (which must then operate at extremely low efficiency), it is better to turn off one of the pumps and increase the efficiency by increasing the pump speed. regulated above the nominal value. This possibility is limited by the suction conditions, which worsen with increasing rotational speed, and by the strength of the pump structure, as well as by the danger of approaching the critical rotational speed with the risk of resonance.

Another recommended option is to regulate not one but all pumps. For example, in the situation shown in Fig. 2 obtaining a capacity of 500 m3/h would require reducing the speed of both pumps operating in parallel to slightly below 1400 rpm, and in such a case both pumps would operate with a capacity of 250 m3/h each, still within the permissible range. An unfavorable feature of such a solution is, of course, the need to purchase an additional inverter, but in return we would obtain a higher operating efficiency of the pumps.

The choice of the optimal solution depends on the required regulation range and on the pump characteristic curve (especially on how flat the efficiency characteristic is). Therefore, in each case, an analysis similar to that shown above should be performed on specific characteristics fig. 2.

In general, it should be stated that regulation by changing the rotational speed is best suited to circulation systems in which the system characteristics are parabolic. It is then possible to select the pump in such a way that it operates with optimal efficiency over the entire rotational speed range (this requires finding a pump for which the parabola of the optimal efficiency points at different rotational speeds coincides with the parabola of the system characteristics). However, in a situation where it is required to maintain a constant lifting height with variable efficiency, regulation by changing the rotational speed is not optimal, because when the speed is reduced, the pump enters the unfavorable range of too low efficiency.

Theoretically, in order to obtain the possibility of a wide change in performance at constant pressure, the optimal solution would be to use positive displacement pumps (piston, plunger, etc.) regulated by changing the rotational speed, because these pumps are able to deliver the required pressure regardless of the performance and without unfavorable movement effects. The barrier to such a solution is the higher price of positive displacement pumps compared to centrifugal pumps with similar parameters.

Summary:

The commonly used solution to obtain variable performance at constant pressure, which involves the use of speed control of one pump operating in parallel with constant speed pumps, is not a technically advantageous solution, as it causes extremely unfavorable operating conditions for the regulated pump within specific control ranges.

Dr. Eng. Grzegorz Pakula

1. Introduction

In a steam power plant, cooling water pumps operate in the turbine condenser cooling circuit, whose task is to reduce the condensate temperature, which helps to obtain a higher vacuum value behind the turbine, thus increasing the overall efficiency of the thermal cycle. There are two basic solutions. If the power plant has an external source of cold water for cooling (e.g. a river, a sea lake), the flow takes place in an open circuit, i.e. the pumps draw cold water from the source, which returns to the source after flowing through the condenser. In such a case, cooling water pumps do not have to have a significant lifting height, because their task is only to overcome the flow resistance in the system. This means that propeller pumps are generally suitable for such applications, as they have significant efficiency at a low lifting height. However, if the power plant does not have an external source of cooling water, the pumps operate in a closed circuit, feeding water heated in the condenser to the cooling tower, where the water is cooled to the initial temperature. In such a system, the cooling water pumps must additionally overcome a geometric lifting height related to the height of the cooling tower. Since the power consumed by a pump is proportional to its lifting height, it is advantageous from an energy point of view to design cooling towers that will provide the required cooling intensity at a minimum height. For this reason, over the years there has been a tendency to build lower and lower cold stores. In currently constructed high-power power units equipped with cooling towers, the lifting height required from pumps usually exceeds 20 m. At the same time, the required efficiency of pumps in units with a power of several hundred megawatts is in the order of several tens of thousands of cubic meters per hour, depending on the power of the unit and the planned to install the number of pumps. This means power consumption of several MW, which is a significant part of the energy consumption for the unit's own needs. With this combination of parameters, diagonal pumps are used as cooling water pumps working with cooling towers.

Due to the above-mentioned significant power consumption, it seems advisable to adjust the pump capacity to the weather conditions, i.e. reduce the cooling water pump capacity in periods when the cooling intensity is higher due to lower air temperature, as well as adapt the cooling water pump capacity to the block load. This fact is not as obvious as it may seem, because reducing pump efficiency, in addition to reducing power consumption, also reduces the cooling intensity of the condenser. By maintaining the pump efficiency at the same level despite lower air temperature or despite lower block load, a greater negative pressure in the turbine condenser can be achieved, and thus increase the power and efficiency of the steam cycle. It should be borne in mind that the range of block power regulation should not be too wide because coal blocks operating at reduced power show reduced efficiency. In a properly configured energy system, coal units with a capacity of 1000 MW should not be used to regulate the system's power, but should operate at full load and with optimal efficiency. However, adapting the power generated in the system to demand should be carried out by peak power plants that are better suited for this purpose, such as gas-fired hydro and thermal power plants. In many coal-fired power plants, despite the existing possibilities of regulating the efficiency of cooling water pumps, they are operated with constant efficiency. Determining whether the capacity control of cooling water pumps, and if so, to what extent, is optimal in terms of energy requires an analysis including the steam cycle, the characteristics of the condenser, the cooling tower and the control characteristics of the pumps. Such an analysis is not the subject of this article, the intention of which is only to provide data for it in the form of information on possible methods of pump regulation. From the point of view of pump regulation, it is important to state that in the case of cooling water pumps for power units, the rational range of regulation is not deep because excessive limitation of the cooling liquid flow brings more losses in the steam cycle of the power unit than benefits from limiting the pump power.

2. Characteristics of the cooling water system and pumps

It should be stated that the question "what is the optimal method of regulating diagonal pumps" is not properly asked, because there is no regulation method that is best in every case. First of all, the pump should never be considered alone but always in cooperation with a specific pumping system. If we intend to analyze the operation of the pump with variable, adjustable capacity, then this capacity corresponds to a certain required lifting height. This height results from the characteristics of the pumping system, because from this characteristic it is possible to read the lifting height required to pump a specific capacity through the system.

There is a significant static head in the cooling tower circuit. It results primarily from the geometric height of the cooling tower to which the pump must supply water, and, moreover, if the cooling tower is equipped with water spray nozzles, a certain pressure is required in the collector supplying these nozzles. In addition to a constant static head, the pump must overcome flow losses in the circulation, which, as we know, increase approximately with the square of the efficiency. Characteristics of the system (fig. 1) is therefore a parabola starting from the vertical coordinate axis at the point corresponding to the static height.

This characteristic for the cooling tower circuit is usually flat, which means that in the operating capacity range, flow losses are significantly lower than the static head. The length of the pipelines is not significant, as it results from the distance of the cooling tower from the condenser, which usually does not exceed several hundred meters, and the diameters of the pipelines should be selected to ensure adequate capacity. The use of pipelines with too small a diameter, which results in an excessive increase in losses, would mean unnecessary expenditure on power consumption needed to overcome excessive resistance to the flow of cooling water throughout the entire period of operation of the unit. It can therefore be assumed that, as in Fig. 1, the system characteristics are flat, i.e. the change in efficiency in the regulation range from minimum to maximum efficiency is accompanied by a slight change in the lifting height. The pump parameters will therefore change according to the section of the system characteristic shown as below Fig. 1 between Qmin and Qmax, which is approximately horizontal, and in any case the changes in head with a change in capacity will not be significant compared to the static head.

When analyzing possible regulation methods, it is necessary to take into account how the pump efficiency and its power consumption will change at points located on the mentioned section of the characteristic curve, which determines the energy savings possible due to regulation. The analysis should also take into account the period of time during the year that the pump will operate at specific points in this part of the system characteristics. Possible energy savings for driving pumps should be compared with the effects of reduced efficiency on the operation of the condenser and the entire circuit.

In order to assess the effects of pump regulation, the course of its power characteristic is important.

For diagonal pumps operating at constant rotational speed, there are two typical cases of this characteristic as in Fig. 2: power consumption decreasing with efficiency (curve A) or a characteristic with a local maximum (curve B), with the efficiency increasing from zero to Q1 power consumption decreases, then increases with a further increase in efficiency from Q1 to Q2, reaches a local maximum at Q2, and decreases again with a further increase in efficiency.

3. Typical adjustment methods

Typical methods used for all pumps as well as methods specific to diagonal pumps can be used to regulate diagonal pumps.

The first group includes:

3.1. Choking
Throttling can be used to limit the pump capacity at constant speed. The advantage of this method is its simplicity, but from the energetic point of view it is not advantageous. If the pump has a power characteristic that decreases with efficiency (curve A in Fig. 2) this throttling is completely pointless as a control method because limiting the efficiency not only does not result in any energy savings but, on the contrary, leads to an increase in power consumption. Some savings can be achieved with pumps with a local maximum curve (curve B in Fig.3) because reducing the efficiency on the section Q2 to Q1 leads to a reduction in power consumption, but the savings are small compared to other methods. Choking is discussed here mainly for the sake of order, as it is unlikely that this method will prove to be optimal as a result of the analysis.

3.2. Discount
Regulation by bleed involves opening the bleed valve located on the discharge pipeline and returning part of the pump's capacity to suction. The pump efficiency then increases due to the fact that the pressure on the discharge side decreases, but less efficiency flows into the discharge pipeline than before the bleeder was opened. The difference between the pump capacity and the flow through the discharge pipeline returns through the bleeder. Energy savings occur when the pump has a decreasing power characteristic, because then when the bleeder is opened, its consumption decreases. Regulation by bleed, similarly to throttling, is relatively simple and inexpensive to implement, but it provides significant savings only for pumps that have a strongly falling power characteristic within the required regulation range.

3.3. Turning on and off pumps working in parallel.
If several pumps operating in parallel are installed, the performance can be regulated in steps by turning the appropriate number of them on and off. This control method is well suited to systems with flat characteristics, such as: Fig. 1 because in such a case, after one of them is turned off, the others continue to operate at a similar lifting height and therefore remain in the high efficiency area. As mentioned, however, this method of capacity regulation is only coarse. In addition, installing more pumps complicates the installation, and replacing a higher-efficiency pump with more lower-efficiency pumps usually reduces their efficiency. However, this is not a significant decrease. For example, diagonal pumps with a capacity of 50 m3/h you can expect an efficiency of 88-90%, while pumps with a capacity of 10 m3/h, an efficiency of 85-87% can be expected.

3.4. Regulation by changing the rotational speed
Regulation by changing the rotational speed for energy efficiency reasons is now widely used in pump technology. However, in its specific application to the regulation of diagonal pumps cooperating with cooling towers, it also has some disadvantages. First of all, for medium-voltage electric drives with powers of several megawatts, the frequency converters required to change the engine speed are characterized by a significant cost. Moreover, their development requires special large-sized rooms with an extensive ventilation system due to the significant heat generation. Moreover, in the case of regulation by changing the rotational speed, the typical "shell" characteristic showing changes in efficiency has the shape as shown in fig. 3. The point of highest efficiency moves along a parabola originating from the origin of the coordinate system, while the area of ​​high efficiency is located near this parabola. This means that in the case of a flat system characteristic such as that shown in Fig. 3, the pump, when reducing efficiency, leaves the area of ​​optimal efficiency.

4. Adjustment methods specific to diagonal pumps

The well-known methods discussed above apply to all pumps. However, for diagonal pumps, additional control methods are known that are not used, e.g. for centrifugal pumps.

4.1 Adjustment by changing the angle of the rotor blades
Pumps are built in which the blades are not permanently attached to the impeller hub, but are mounted on rotating pins, thanks to which the angle of their setting can be changed. Angle setting is usually done by a mechanism inside the impeller hub and hollow shaft and can be made while the pump is moving. By analogy to water turbines of a similar design, such pumps are sometimes referred to as "Deriaz pumps". This method of regulation is effective, but apart from this advantage, this solution also has disadvantages. The design of the pump, compared to a typical diagonal pump, is much more complicated, which means that the prices of pumps of this type are much higher than in the case of diagonal pumps with fixed blades. In order for the blades rotating around their axis not to create an excessive gap in relation to the hub and impeller cover, the hub and cover must have a spherical shape, which is not advantageous from the point of view of flow hydraulics and causes pumps with adjustable blades to achieve lower efficiencies compared to diagonal pumps. with fixed blades. The presence of a gap between the blade and the hub causes additional gap flow losses. During operation, the problem is maintaining the correct initial gap size. Its enlargement as a result of flushing by slot flows leads to an increase in hydraulic losses, while the formation of deposits on the hub hinders the movement of the blades relative to it. The complexity of the design leads to a potential reduction in pump reliability. The blades of diagonal pumps, due to their large surface, are subjected to significant loads from hydraulic pressure forces, which causes a significant load on the blade hub and other elements of the angle setting mechanism. As a consequence, this leads to the formation of backlashes in this mechanism and to an increase in the pump vibration level.

4.2 Adjustment using the pre-rotation vane

A less known method of adjusting diagonal pumps is the use of a pre-rotational guide vane. Such a solution is shown in fig. 4. It deserves more attention and will therefore be discussed in more detail here.

Fig. 4. Pre-rotational handlebar

The principle of operation of such a steering wheel results directly from the basic equation of turbomachinery, the so-called Euler's equation, which gives the approximate dependence of the pump lift height on the velocity components in the inlet and outlet cross-sections of the pump blades:

H = (u2 cu2 – u1 cu1 ) / g, (1)

where g is the acceleration due to gravity, u1 and u2 are the absolute peripheral velocities of the blade at the inlet and outlet resulting from rotation, and cu1 and cu2 are the components in the circumferential direction of the absolute velocity of the liquid at the inlet and outlet of the blade. (The above formula strictly applies to a palisade with an infinite number of blades, but it is useful for a qualitative analysis of the phenomenon also for rotors with a finite number of blades).

In a typical pump, the liquid flow into the impeller takes place along the axis, which means that the projection of the absolute speed on the circumferential direction cu1 is equal to zero.

In this case, equation (1) reduces to a simplified form:

H = u2 cu2 / g, (2)

If we use a preliminary guide as in fig 4, then we can introduce prerotation, as a result of which in the formula (1) the second term takes on a value other than zero and influences the lifting height achieved by the pump. If the primary guide directs the liquid in the direction of rotation of the impeller, the cu1 component is positive and the second component in the formula (1) lowers the lifting height. Figuratively, this effect can be explained by the fact that the liquid "escapes" from the rotor blades, as a result of which the rotor transfers less energy to it. If the prerotation is in the opposite direction to the rotation of the impeller, the pump lifting height is increased.

At the test station of the Powen-Wafapomp SA Group, model tests were carried out on the regulation of the 200D40 pump, intended as a cooling water pump for 1000 MW units, using a primary guide. At a rotational speed of 375 rpm, the pump has nominal parameters Q = 43 m750/h and H = 3 m. Long-term tests of a pump with such high efficiency on a test stand are impossible, therefore, in order to examine the regulatory characteristics, a model pump geometrically similar to the main pump, reduced in size, was built. on a scale of 24.5 : 1. The model pump had the following parameters: capacity Q = 5,4 m1100/h, lifting height H = 3 m at a rotational speed of n = 13 rpm. The photo of the model pump is shown as figure 6.

Fig.5. Dimensionless control characteristics

Results obtained on a model pump (Fig. 5) are presented in dimensionless form, relating them to the values ​​of efficiency, lifting height and efficiency at the nominal point. This allows predicting the characteristics of all pumps that are geometrically similar and have the same speed, regardless of their scale. On Fig. 5 The H(Q) characteristics for different angles of the pre-rotation guide vane are shown. An angle of 90° means setting the chord of the blades along the pipeline axis, i.e. no prerotation, while decreasing the angle means introducing prerotation consistent with the direction of rotation of the impeller, and therefore, as written above, lowering the pump parameters.

Analysis of the results shows that the recommended regulation range is narrower than in the case of pumps with adjustable blades. However, it should be borne in mind that, as stated in the introduction, the required range of regulation of cooling water pumps in the case of power units is not wide for thermodynamic reasons. Regulation using a pre-rotation guide is therefore an interesting alternative to pumps with adjustable vanes, as it allows you to achieve the expected effect with a much less expensive, simpler, and therefore more reliable solution. The primary director has a low investment cost, comparable to the cost of a throttle valve and many times lower than a frequency converter or a pump blade angle setting mechanism. It is worth emphasizing that the mechanism for setting the angle of the primary guide vanes is completely independent of the pump, which means that any problems with its functioning do not cause problems with the pump's movement, which occurs in the event of a failure of the pump mechanism with adjustable vanes.

Fig.6. A model pump for testing a pre-rotational steering wheel.

5. Summary

The method of regulating the parameters of cooling water pumps for power units should not be selected beforehand but should result from an analysis that takes into account both the thermodynamic aspects of the power plant's steam cycle and the characteristics of the pump and pumping system. After initially determining in what capacity range the regulation of cooling water pumps is beneficial from the thermodynamic point of view, this range should be analyzed in the appropriate section of the pump system's characteristic curve, taking into account the statistically expected number of hours during which the system will operate at a specific point of the characteristic curve. Then, it is necessary to estimate what energy savings can be achieved by using various control methods to obtain such specific points on the system characteristics.

The article discusses various control methods, focusing on methods specific to diagonal pumps. Attention was paid to the possibility of adjustment by using an initial pre-rotation vane, which is not widely known, but deserves attention because it allows to obtain interesting effects using a low-cost and simple, and therefore reliable, design solution. This solution is particularly suitable for use in situations where the required range of capacity regulation is not wide, as is the case with cooling water pumps for power units.

Authors:

• Dr. Eng. Grzegorz Pakuła is Technical Director, member of the management board POWEN SA,

• Wiktor Piasecki, M.Eng is the Chief Specialist for Diagonal and Propeller Pumps in the Design Office POWEN-WAFAPOMP SA Group

• Artur Szarszewski, M.Eng is a graduate of the MEIL Faculty of Warsaw University of Technology, works at the Test Station in POWEN-WAFAPOMP SA Group.

Literatura:

1. A.Troskolański, Sz.Łazarkiewicz, Centrifugal pumps, WNT, Warsaw. 1973         
2. W. Jędral, Centrifugal pumps, PWN Scientific Publishing House, Warsaw 2001,
3. Sz.Łazarkiewicz, A.Troskolański."Modern directions in the design of centrifugal pumps" WNT, Warsaw 1966  

1. Introduction

The primary function of the mine drainage system is to ensure safe mine operation by eliminating water hazards. Achieving this goal requires incurring costs that are important for the profitability of the mining plant. These costs should be controlled and optimized so that they deviate as little as possible from the inevitable minimum resulting from the laws of physics.

2. Drainage cost structure

In order to optimize dewatering costs, it is advisable to estimate the proportions of their individual components to each other. We will estimate the main components of dewatering costs below on the example of a main dewatering pump with typical parameters: efficiency Q = 500 m³/ h, the hight of raising H = 800 m.

The estimates below do not refer to any specific pump type and are intended only to establish the order of magnitude of the individual components. Drainage costs include:

a) Investment costs, i.e. the costs of purchasing and installing the pump unit. For a pump with the parameters given above, these costs amount to PLN 600. zloty. (the exact value depends on the material, manufacturer, etc.). If we assume that such a pumping unit is depreciated over a period of 10 years, then the investment cost will appear in the annual dewatering costs in the form of depreciation at the level of PLN 60. zloty. In practice, main drainage pumps are in operation for longer than 10 years, which makes the investment cost even less important.

b) Electricity costs to drive the pump. The pump with the parameters given in this example consumes a power of 1.5 MW. A typical main drainage pump operates approximately 10 hours a day, or 3650 hours a year. This gives an annual energy consumption of 5475 MWh. Assuming the price of electricity is PLN 300/MWh, we obtain the cost of annual energy consumption by the pump at 1. PLN 6425 million.

c) Cost of renovations. The scope and frequency of renovations depend on the quality of mine water and the design and level of workmanship of the pump. Based on experience, it can be assumed that on average the pump requires a major overhaul approximately every 10 hours of operation, i.e. approximately once every three years. The cost of such a renovation can be estimated at approximately PLN 000. PLN, so the annual share of renovation costs in drainage costs is approximately PLN 150. zloty.

d) Service costs. Modern pumps can operate unattended, in automatic mode, requiring only periodic inspection. The cost of servicing therefore results more from the regulations and policies of a given mining plant regarding safety and supervision of the main drainage system than from technical reasons. Let us assume that the annual cost of servicing the pump corresponds to the cost of one full-time job, and is therefore at the level of PLN 50. zloty.

The following conclusions follow from the above estimation:

• The total cost of dewatering using a typical main dewatering pump (approx. PLN 1.8 million) is dominated by the cost of energy (PLN 1.64 million), constituting over 90% of the total costs. Reducing drainage costs must therefore focus on reducing energy consumption.

• The cost of energy consumed depends strongly on the energy efficiency of the pumps. This efficiency, in turn, depends on the quality of the pump's design and workmanship. In difficult mine conditions, the initial efficiency of the pump unit provided by the manufacturer changes, and therefore the cost of energy consumed depends in the long run on the applied maintenance policy, which determines the average efficiency of the pumps over the period of operation.

The proportions between the individual components of drainage costs are shown in fig.1.

Fig. 1. Proportions between individual components of drainage costs

3. The impact of renovation policy on the cost of main drainage.

Staying with the above example of the main drainage pump, we will analyze the relationship between the cost of energy to drive the pump, which dominates the drainage costs, and the renovation policy.

The power consumed by the pump can be calculated from the formula:

N = = γ QH/ η_z ,

where η_z means the efficiency of the pump unit, which is the product of the efficiency of the pump and the engine.

We are considering a main drainage pump with a capacity of 500 mXNUMX³/h and a lifting height of 800 m pumping water with a density of 1000 kg/m³. For the above parameters, the efficiency of the pump unit should be at least 75%, at which the power consumption will be 1453 kW. However, if the efficiency of a set with such parameters was 70%, the power consumption increased to 1557 kW, and for an efficiency of 65% to 1676 kW. As you can see, with the above parameters, such a drop in efficiency means an increase in power consumption of 200 kW. If the pump runs 10 hours a day, i.e. 3650 hours a year, this corresponds to an increase in energy consumption of 730 kWh. If we assume the price of a kilowatt hour is PLN 000, the cost of additional energy will be PLN 0.3. zloty. annually. The difference in the annual energy cost resulting from differences in the quality of the renovation is therefore at the level of the renovation price, and definitely exceeds the differences in renovation prices offered by individual contractors. The comparison of the above numbers leads to the conclusion that selecting contractors for pump renovations solely on the basis of the renovation price without paying attention to its effect in the form of efficiency is economically unjustified, because the slight savings obtained on the renovation price may be far lower than the losses on additional energy consumption. The correct approach is therefore to compare prices with the specified renovation quality required from the contractor, expressed in the energy efficiency achieved after the renovation. The condition for enforcing such requirements is to carry out measurements of the pump parameters after the renovation, confirming that the required efficiency has been achieved. It should be noted that the mechanical efficiency of the pump itself (smooth operation, good overrun, appropriate vibration level, etc.) is not an appropriate measure of the quality of the renovation. Therefore, the criterion for accepting the pump after renovation should be, in addition to the required basic parameters (efficiency, lifting height), measured energy efficiency. However, when formulating the requirement in this respect, one should be realistic, because in the case of a heavily used pump, it is not always possible to achieve the efficiency of a new pump after renovation. Nevertheless, the permissible deterioration in the efficiency of the pump after renovation in relation to the efficiency of the brand new pump cannot be too significant (the acceptable value of efficiency reduction should not exceed several percent, and most often should not exceed 219-2%), otherwise, due to a significant increase in the cost of energy consumed, the economic sense of renovation becomes questionable and purchasing a new pump becomes a real option.

Another conclusion resulting from the comparison of renovation prices and energy prices to drive the pump is that there is a certain optimal frequency of major renovations. In the period between subsequent major overhauls, the pump's efficiency gradually decreases due to wear. If renovation is delayed until a failure occurs or until the pump is no longer able to provide the required performance, the average efficiency over the period of operation decreases and the average cost of pumping a cubic meter of water increases. Increasing the frequency of renovations to restore pump efficiency increases the renovation component of drainage costs, but reduces energy costs. By monitoring and analyzing both of these operating cost components on an ongoing basis, it is possible to determine the optimal frequency of renovations, allowing for maintaining the drainage cost index discussed below at a minimum level.

4. Dewatering energy consumption index

From the point of view of energy efficiency, what is important is not the instantaneous power consumption but the amount of energy used to pump out 1 mXNUMX.³. By dividing the power consumption expressed in kW by the efficiency expressed in m³/h we get an indicator showing how many kilowatt hours are needed to pump out one cubic meter of water in a specific pumping system. Power consumption is proportional to parameters and efficiency Q and lifting height H. There is a minimum amount of energy that must be used to pump a liquid to a height H, below which you cannot go.

This minimum energy consumption for pumping a cubic meter is:

N_min /Q = γH_g,

where H_g is the geometric height to which water is pumped.

The energy consumption coefficient determined in this way cannot be used to compare different drainage systems, because it depends on the geometric lifting height, which differs for individual systems. However, this indicator allows you to assess the energy efficiency of a given drainage system by comparing the actual energy consumption for pumping a cubic meter with the minimum consumption.

Going below this minimum specific energy consumption is physically impossible. During operation, efforts should be made to ensure that the actual specific energy consumption exceeds the above minimum as much as possible.

As the two formulas above show, increasing the actual energy consumption above the physical minimum results mainly from two reasons:

a)   From the fact that the actual efficiency of the pump unit η_z is less than unity. The value of the energy efficiency of the pump unit is mainly influenced by the renovation policy, referred to in the previous point, and the selection of the pump for the system, as referred to in the next point.

b)   This is because the total head H in the pump system is greater than the geometric head by the amount of losses in the pipelines.

While the geometric height is a constant value, the amount of losses can be influenced to some extent. The main drainage system usually consists of several pipelines, often of different diameters. In order to reduce losses, it is advisable to pump through all available pipelines simultaneously because then the flow rates and therefore losses are reduced compared to pumping through a single pipeline. For the same reason, it is energetically advantageous to pump as few pumps as possible at the same time. In the drainage system, sources of additional losses should be eliminated, e.g. in the form of smaller diameter pipeline sections, fittings with too high resistance coefficients, etc. For example: in a pipeline with a diameter of 400 mm, with a flow of 1000 m³/h the loss amount is approximately 1.2 m per 100 m of pipeline length. With the same capacity in a pipeline with a diameter of 300 mm, the losses increase to approximately 5.2 m per 100 m of length. Eliminating this type of loss requires some investment. Some energy savings can also be achieved without any investment by ensuring the proper quality of system operation, e.g. by ensuring that all cut-off fittings along the pipeline route are fully open during pumping, that the suction strainer is not contaminated, etc. If this does not threaten the safety of the pumping station due to reduced retention it is advisable to avoid pumping when the liquid level in the suction tank is low, i.e. turn off the pumps until the water flows to a higher level in the tank.

It should be emphasized that actions aimed at saving energy by reducing flow resistance only produce the best results when they go hand in hand with adjusting pump parameters. If we take action to reduce resistance, e.g. by pumping through more than one pipeline or limiting resistance on fittings, it may turn out that due to changes in the characteristics of the system, the pump operating point will shift to higher efficiencies, which will limit energy savings as a result of operation. pumps outside the optimal efficiency range. Therefore, if significant reductions in flow resistance are achieved, in order to fully utilize them, this should be accompanied by a corresponding reduction in pump parameters.

5. The influence of the selection of a pump for the system on efficiency.

It should be emphasized that the most important influence on the value of energy consumption is not the maximum (catalog) efficiency value, but the efficiency at the actual operating point. As we know, the pump operating point occurs at the capacity at which the pump characteristics intersect with the system characteristics. In order to select the pump correctly, it is necessary to estimate the characteristics of the system and therefore determine how the losses change with the efficiency. The common practice of selecting a pump with maximum diameter impellers, with the closest number of stages at which the head exceeds the required one, in some cases leads to installing a pump with a head with an excess of several dozen meters, which causes the pump to operate in the system with too high efficiency. This is unfavorable in terms of energy for two reasons: firstly, for increased efficiency, the pump efficiency is reduced compared to the maximum, and secondly, increased efficiency causes increased flow resistance. Excessive increase in efficiency leads, in extreme cases, to engine overload and cavitation. Closing the discharge valve avoids these phenomena, but means the pump operates with increased specific energy consumption due to the need to overcome additional losses on the valve. You should be aware that closing the discharge valve causes a decrease in absolute power consumption, but most often means an increase in specific power consumption related to the pumped cubic meter. Operating practice is often to pump with fully open discharge valves until there are no problems with engine overload or cavitation, and if these problems do occur, they are eliminated by throttling the pump. This method of operation ensures the avoidance of basic traffic problems, but does not guarantee that pumping is carried out optimally from the energy point of view. In many cases, the optimal rate of energy consumption per cubic meter can be obtained by limiting the pump parameters by reducing the diameter of the impellers. This solution avoids throttling with the valve.

In order to determine the best choice of pump for the system in terms of energy, it is necessary to analyze the characteristics of the pump and the system. In principle, this type of analysis should be carried out at the stage of design work aimed at selecting a pump for the system. In practice, it is advisable to carry out such an analysis during operation, because the actual characteristics of a specific pump and the actual characteristics of the system often differ from the theoretical characteristics analyzed at the design stage. For this reason, during operation, it is advisable to verify the selection of the pump for the system in order to ensure that the pump operates at its highest efficiency level.

Main drainage pumps have a high lifting height from the stage and therefore do not densely cover the entire range of possible lifting heads. It may happen that a pump with a certain number of stages does not provide the required lifting height, and a pump with one more stage number has excess parameters. Such a case is shown in fig. 2. This is an example case. The pump and pipeline characteristics do not refer to any specific pump type or system but are drawn for illustrative purposes only. In this case, a pump with a capacity of 500 m should be selected³/h to the pump system, the characteristics of which are shown in Fig. 2 with a dotted line.

This characteristic corresponds to the characteristics of a pumping system with a geometric height of 710 m and a pipeline with a diameter of 300 mm and a length of 800 m. For a capacity of 500 m³/h, the height of losses is approximately 10 m, i.e. the pump with this capacity should be selected for a lifting height of 720 m. We are analyzing the selection of a pump with a nominal capacity of 500 m for this system.³/h (i.e. appropriate for the required capacity) and a height from a step of 70 m. The H(Q) characteristics for this pump are shown in Fig. 2 with a solid line, and the efficiency characteristic η(Q) with a dashed line. The 10-stage pump with its nominal capacity has a lifting height of 700 m, so it is 720 m short of the required 20 m. As a result, the characteristics of the 10-stage pump intersect with the characteristics of the pump system with a capacity of approx. 430 m³/hi at a head of approx. 718 m. This pump therefore does not meet the requirements. Not only does it not provide the required performance, but its operating point is outside the optimal efficiency range. As seen from Fig. 2, the pump whose highest efficiency is approximately 79%, in this system, due to incorrect selection ("choking" by excessive lifting height), would operate with an efficiency of approximately 74%. The power consumption (which can be calculated from Q, H and efficiency or read from the N(Q) characteristic not shown in the figure) would be approximately 1137 kW in this case, assuming that clean water with typical specific gravity is pumped.

From the point of view of energy efficiency, what is important is not the instantaneous power consumption but the amount of energy used to pump out 1 mXNUMX.³. By dividing the power consumption expressed in kW by the efficiency expressed in m³/h we get an indicator showing how many kilowatt hours are needed to pump out one cubic meter of water in a specific pumping system. In the above case, for a 10-stage pump, we obtain an indicator of 2.64 kWh/m³. It should be noted that the value of this indicator in this case is positively influenced by the fact that with reduced efficiency there is a reduced head due to lower flow losses in the pipeline, while a negative impact is caused by reduced efficiency due to the pump operating outside the optimal range.

In turn, if an eleven-stage pump was used (with an excess head above the requirements of this system), its operating point, resulting from the intersection of the pump characteristics with the pipeline characteristics, would be at a capacity of approx. 645 m³/hi at a lifting height of approximately 728 m. The pump should not operate at such a point. First, it would be outside the optimal efficiency range, as the efficiency at this operating point is only 71%. Moreover, as follows from the characteristics, NPSH is required_r at this point it increases to approximately 7 m, which significantly limits the pump's suction capacity and may result in cavitation if the water level in the suction tank slightly decreases. The power consumption of the pump would be approximately 1802 kW, which is more than the power of the engine selected for nominal parameters. If you consciously decide on such a selection, you could avoid overload by installing a motor with increased power, but this would be unfavorable from the energy point of view, because the kilowatt-hours per pumped cubic meter would in this case amount to 2.79 kWh/m³. The increase compared to the previously calculated value is due to two reasons: firstly, due to the increase in efficiency, the amount of losses in the pipeline increases to approximately 18 m, and secondly, the pump operates with even lower efficiency.

To improve the selection quality, it is advisable to use rotors with a reduced diameter. This diameter should be selected so that the pump capacity is close to the required 500 m³/ h.

It is advisable to use a diameter that gives slightly higher efficiency, which is due to two reasons.

Firstly, it provides a certain reserve of pump parameters for wear during operation, and secondly, as the degree of reduction of the impeller diameter increases, the efficiency drop deepens, so the diameter should be reduced to the smallest possible extent. On Fig. 2 an exemplary characteristic of an 11-stage pump with reduced diameter rotors is shown with a dash-dot line, which intersects with the system characteristic at 520 m³/h and lifting height of approx. 721 m. Due to the reduction of the rotor diameter, the efficiency characteristics η(Q) also change. As is known, the maximum efficiency decreases and its position shifts towards lower efficiency. The efficiency characteristics of a pump with reduced impeller diameter are shown in Fig. 2 with a dashed line with double points. How does it show the efficiency of the pump at 520 m?³/h will be approximately 76%. This, in turn, results in a power consumption of 1344 kW and an energy consumption index of 2.585 kWh/m³. As you can see, this indicator is more favorable than for 10- and 11-stage pumps with full rotor diameters. The degree of rotor diameter reduction should be optimized in such a way as to obtain the most favorable energy consumption rate possible. As can be seen from the above, an excessive value of the diameter leads to a reduction in efficiency as a result of operating at too high a capacity.

On the other hand, too deep a diameter reduction also causes a decrease in efficiency. It is therefore recommended to analyze several degrees of the rotor diameter in order to determine the value ensuring the optimal energy consumption rate for pumping a cubic meter. In addition, as mentioned, one should also take into account the reserve of parameters for pump wear, as well as a very important factor, which is NPSH in terms of efficiency at the operating point.

The example of pump selection shown above illustrates the statement that selecting a pump for a specific, single point Q, H is not a correct practice, as it does not enable a full analysis of the effects resulting from the mismatch of the pump parameters to the system. Only the analysis carried out on the characteristics of the pump system allows predicting the resulting operating point with all the consequences in the form of efficiency, power consumption and suction capacity. As mentioned, this example does not concern any specific case, but qualitatively illustrates the impact of pump selection on the achieved efficiency.

Fig. 2. Selection of the pump for the system.

6. The influence of the method of operation on efficiency.

As mentioned above, maintaining high, average efficiency during operation depends on the renovation policy. However, the pump's efficiency is also influenced by the way it is operated between renovations. In addition to such obvious issues as avoiding too high mechanical resistance due to lack of lubricant in the bearings or too tight tightening of the stuffing boxes, the operation of the relief disc is also important.

Pump efficiency depends largely on the relative position of the impellers and guide vanes. This position in a pump with a relief disc depends on the degree of wear. If the rotating assembly moves so much due to the wear of the unloading disc that the streams flowing from the rotors hit not the guide vane channels but their walls, the efficiency deteriorates significantly. The pump efficiency is the highest when the axes of the rotor channels and guide vanes coincide. The asymmetrical positioning of the rotor and vane channels relative to each other results in reduced efficiency even if the stream does not hit the wall yet. Therefore, in order to maintain high operational efficiency, the staff should monitor the position of the shaft travel indicator and strive to ensure that the rotating assembly is in the optimal position. This can be achieved by periodically replacing the relief disc rings and/or shims in front of the disc.

SUMMARY AND CONCLUSIONS

The main component of the cost of main drainage is the cost of electricity, which may exceed 90% of the total cost.

In order to reduce it, it is necessary to conduct an appropriate renovation policy consisting of:

a) carrying out major overhauls with the optimal frequency ensuring the maintenance of high average efficiency during the period of operation,

b) requiring contractors performing major overhauls of pumps to obtain guaranteed high energy efficiency and checking this condition at a testing station as a criterion for accepting the overhaul.

The correct selection of pumps for a given pumping system is also very important and should be carried out individually for each case.

The pump should also be operated in an appropriate manner to maintain high energy efficiency.

In order to optimize the cost of pumping, it is advisable to record the amount of water pumped and the amount of electricity used for this purpose, and to monitor and analyze the energy consumption index calculated on this basis.

Dr. Eng. Grzegorz Pakula