## Abstract

Reduction of pressure at pump houses are the most feasible and most advantageous as the pipe design method that is used, is more than 200 years old and is based up on 19th century production technology. My research focuses on a general approach on improving and reducing the pressure loss of these pipe elements with the help of non-conventional methods, thus resulting in a lower and more optimal energy usage of pump houses. The problematic zones are identified with the help of numerical modelling, geometry changes can be made and tested the same method. The geometrical changes aiming at pressure loss reduction follow non-conventional ideas, form hemodynamic and other biomechanics sources. Pipes in the pump houses are designed, for more than 50 years of operation. Even a small pressure loss reduction with this new method will mean large amount of energy saving in total. The results show that 15%–60% of pressure loss reduction is feasible, according to the complexity of the geometry. Pressure loss reduction will reduce energy consumption of water pumps which will result in a more efficient water works operation.

## INTRODUCTION

Pump houses have the highest velocity values of the water network system and the highest pressure loss values are present here also. Reduction of pressure at these facilities are the most feasible and most advantageous as the pipe design method that is used in pump houses, is more than 200 years old and is based up on 19th century production technology. The basic standards for pipe elements (elbow, T-junction, collector pipe, etc.) did not change in the past century. The flow characteristics and the pressure loss of these elements are well known in all velocity ranges. During the current design and renovation of these facilities the pressure loss of the pipes are calculated, pumps, and operation methods are chosen accordingly. So we can state that until this point there were no general attempts made to improve these simple but very old pipe element designs, although manufacturing technology has drastically improved and more complex geometry forms can be manufactured economically. My research focuses on a general approach on improving and reducing the pressure loss of these pipe elements with the help of non-conventional biomechanical methods, like hemodynamic, thus resulting in a lower and more optimal energy usage of pump houses.

## METHODOLOGY & RESEARCH

The general structure of the analysis used the following method. Numerical calculations were carried out on the original standard pipe elements with various input velocity range to determine the problematic fluid dynamic zones where pressure loss occurs. The analysis also provided precise pressure loss values of the geometries. The two best indicator quantities of these zones are turbulence kinetic energy and streamlines of the section. The problematic zones have a high turbulence kinetic energy value and the streamlines in and near these zones are not parallel to each other. After the problematic zones are identified geometry changes can be made and tested, these changes do not alter the space required for the specific pipe element. Easy installation and manufacturability are also taken in to account. The geometrical changes aiming at pressure loss reduction follow non-conventional ideas, form hemodynamic and other biomechanics sources (asymmetrical cross section areas, surface modification, complex 3D surfaces etc.). Each idea than was analysed with numerical calculation and the best one with the highest pressure loss reduction was selected.

## RESULTS

The research is still ongoing as there are numerous standard pipe elements in use. Research for new 90° degree pipe elbow and a new geometry for a three inlet collector pipe are complete. For other standard pipe element the research is still in progress, but the results are quite promising, with significant amount of pressure loss reduction.

### Standard 90° pipe elbow with R1,5 radius

This is one of the most common pipe elements used in pump houses. The flow dynamics of this element is extensively researched and documented (Beij 1938). It is know, that the main reason for pressure loss is caused by the formation of two secondary counter rotating vortexes after the elbow (Beij 1938) (Figure 1). This phenomenon is caused by the forced direction change of the water molecules. The starting point of the vortex is located at the beginning of the longer arc of the elbow.

The other cause of the pressure loss at high Reynolds number is the backflow after the short arc of the elbow (Bobek 2014). The back flow and the vortex formation can be decreased by increasing the radius of the elbow, but this is not feasible when the installation space is limited. The new suggested geometry reduces the secondary vortex formations via flaps installed at key areas on the inner surface of the pipe (Figure 2). This reduces and delays the formation of the counter rotating vortexes.

The space requirement for the pipe section remains the same for the new geometry, only the inside of the pipe is altered. This geometrical change is simple, yet cost effective and also could be easily manufactured. The new suggested geometry reduces the secondary vortex formations via flaps installed at key areas on the inner surface of the pipe, this also reduces pressure loss (Figure 3).

### Standard collector pipe with three inlets

Standard collector pipe geometry was analyzed with three DN100 inlet pipes ends and a standard DN200 main pipe. The inlet pipes are perpendicular to the collector pipe and they are connected directly. The collector pipe in our case is followed by a standard 90° pipe elbow The standard geometry has no diameter change element (Figure 4).

This type of collector pipe is mainly used for a two pump operation where one pump is used as a reserve. The typical operation uses two pump and one is shut down. One pump operation also exists at low water consumption periods. Altogether five operation methods were analyzed with inlet velocity range from 0.5 to 3.2 [m/s] (Table 1).

. | Inlet 1 . | Inlet 2 . | Inlet 3 . |
---|---|---|---|

Operation 1 | On | On | Off |

Operation 2 | Off | On | On |

Operation 3 | On | Off | Off |

Operation 4 | Off | On | Off |

Operation 5 | Off | Off | On |

. | Inlet 1 . | Inlet 2 . | Inlet 3 . |
---|---|---|---|

Operation 1 | On | On | Off |

Operation 2 | Off | On | On |

Operation 3 | On | Off | Off |

Operation 4 | Off | On | Off |

Operation 5 | Off | Off | On |

First the original geometry was analyzed. The numerical analyses (Tu *et al*., 2008) showed a large zone where turbulence kinetic energy is high, this is the main source of the pressure loss (Figure 5). The high turbulence in this zone is caused by the perpendicular water inlet, the main flow simply hits the side of the collector pipe.

Three approaches were investigated to minimize the pressure loss, and altogether five geometries were analyzed (Table 2). The first approach used modified standard elements with offset. The second approach integrated guide flaps to the previous geometries. The third approach was based upon the shape of the human Aorta (Ng & Luo 2016).

. | Base . | Geometry V1 . | Geometry V2 . | Geometry V3 . | Geometry V4 . | Geometry V5 . | |||||
---|---|---|---|---|---|---|---|---|---|---|---|

. | geometry Press loss . | Press loss . | Saving [%] . | Press loss . | Saving [%] . | Press loss . | Saving [%] . | Press loss . | Saving [%] . | Press loss . | Saving [%] . |

Operation 1 | 4622.5 | 2886.4 | 37.56 | 4089.8 | 11.52 | 740.21 | 83.99 | 1832.2 | 60.36 | 2835.5 | 38.66 |

Operation 2 | 5157.3 | 1476.9 | 71.36 | 1,749 | 66.09 | 4698.1 | 8.90 | 3747.3 | 27.34 | 1598.9 | 69.00 |

Operation 3 | 3740.1 | 2118.6 | 43.35 | 2962.9 | 20.78 | 3224.1 | 13.80 | 2918.6 | 21.96 | 2051.1 | 45.16 |

Operation 4 | 4231.2 | 2078.7 | 50.87 | 2,396 | 43.37 | 5971.7 | −41.13 | 4401.6 | −4.03 | 2061.6 | 51.28 |

Operation 5 | 4370.5 | 3080.3 | 29.52 | 2394.1 | 45.22 | 2242.7 | 48.69 | 2207.9 | 49.48 | 3087.7 | 29.35 |

. | Base . | Geometry V1 . | Geometry V2 . | Geometry V3 . | Geometry V4 . | Geometry V5 . | |||||
---|---|---|---|---|---|---|---|---|---|---|---|

. | geometry Press loss . | Press loss . | Saving [%] . | Press loss . | Saving [%] . | Press loss . | Saving [%] . | Press loss . | Saving [%] . | Press loss . | Saving [%] . |

Operation 1 | 4622.5 | 2886.4 | 37.56 | 4089.8 | 11.52 | 740.21 | 83.99 | 1832.2 | 60.36 | 2835.5 | 38.66 |

Operation 2 | 5157.3 | 1476.9 | 71.36 | 1,749 | 66.09 | 4698.1 | 8.90 | 3747.3 | 27.34 | 1598.9 | 69.00 |

Operation 3 | 3740.1 | 2118.6 | 43.35 | 2962.9 | 20.78 | 3224.1 | 13.80 | 2918.6 | 21.96 | 2051.1 | 45.16 |

Operation 4 | 4231.2 | 2078.7 | 50.87 | 2,396 | 43.37 | 5971.7 | −41.13 | 4401.6 | −4.03 | 2061.6 | 51.28 |

Operation 5 | 4370.5 | 3080.3 | 29.52 | 2394.1 | 45.22 | 2242.7 | 48.69 | 2207.9 | 49.48 | 3087.7 | 29.35 |

It is visible from Table 2 that the different geometry versions, had strength in different operation conditions. For example the third geometry version is best for two pump operation when the first two pumps are active. The pressure loss in this case is reduced by 83.99%. This geometry uses asymmetrical and ellipsoid profiles based on Aorta shape (Ng & Luo 2016) this reduces the vortex formation and the back flow (Figure 6).

Although this this geometry is ideal for reducing the pressure loss for operation 1, it is not an all-around ideal geometry for all the possible operation combination. Pressure loss is increased more than 40% at operation 4 (only one pump operating at inlet 2) compared to the original geometry (Figures 7 and 8).

The goal of the analysis was to find an all-around optimal geometry fall all operation from a velocity range from 0.5 to 3.2 [m/s]. The final geometry version (V5) met all the criteria and reduced the pressure lass at the full velocity rang and at all of the operation methods (Figure 9).

Geometry V5 has significant pressure loss reduction even at low velocities, the average pressure loss reduction is above 30%.

## CONCLUSION

Pipes in the pump houses are designed, for more than 50 years of operation. Even a small pressure loss reduction with this new method will mean large amount of energy saving in total. The results show that based upon the complexity of the pipe geometry 50–60% of pressure loss reduction can be achieved. The research also showed that it is more efficient to view the pump house pipe network as one geometry and make modifications simultaneously, rather than focusing on the individual pipe elements. This is because the pipe elements placed closed to each other influence each other's flow characteristics. The experience gained also show that comparing streamlines and turbulence kinetic energy zones are the best way to analyze and quantify the problematic zones, that cause the pressure loss. The pressure loss reduction will reduce energy consumption of water pumps which will result in a more efficient water works operation.