The Standard Module - Ventilation Systems

Transcription

1 Chapter 3 The Standard Module - Ventilation Systems 3.1 Introduction The PIPENET Vision Standard module is a tool for steady state flow modelling of networks of pipes and ducts. It can model incompressible and compressible flow networks. It is widely used for modelling ventilation systems in the nuclear and other industries. Such calculations are central to the design process in some industries because good design of ventilation systems is essential for safety. Two examples are considered in this section of the document. A simple extraction system. Balancing a system which has fans on the inlet and extract sides of the system. The first of the above examples will be covered in more detail with several figures of dialog boxes. The other examples are equally important but will show fewer dialog boxes 1

2 3.2 Example 1: Machine shop air extraction system Objectives In this example, we will look at an extract system of the type you might get in a machine shop. The objective of the exercise is to take a system that has been manually designed such that all the duct sizes and fan curves are known. We wish to verify that the system would work as required Initialisation of data The network The network as it is drawn in PIPENET Vision is shown in figure 3.1. Figure 3.1: Network for example 9, a machine shop air extraction system. Options Standard Options Select the Colebrook-White formula for the pressure model, and select proceed with calculation for the warnings control. Units The units used are user-defined, as shown in table 3.1. Fluid This is input through the Init Fluids option. We wish to use air at 15 C, and this is done by entering data into the dialog box shown in figure 3.2(a). Defaults Setting default values can save input time, and are pre-entered into the dialog box shown in figure 3.2(b). These values can be changed for individual items, and it is possible to go back to change default values after the network has been partially input. Any part of the network which is input afterwards will have the new default values. Here we use a default roughness of mm. Property Unit Length m Diameter mm Pressure in H 2 O gauge Temperature C Velocity m/s Flow rate type Volumetric Flow rate m 3 /s Density kg/m 3 Viscosity Pa s Table 3.1: User-defined units for example 9. 2 SSL/TM/0001/01 - c 2006 Sunrise Systems Limited

3 (a) Fluid to be used within our ventilation system. (b) Default values for the ventilation system. Figure 3.2: Fluid (above) and defaults (below) dialog boxes. 3 SSL/TM/0001/01 - c 2006 Sunrise Systems Limited

4 Pipe Type This is generally only used when pipes are to be sized. For the purpose of this example, we skip this section. Libraries Fittings This dialog box (figure 3.3) is used when it is desired to remove certain fittings from the library during the problem input. This would avoid having to scroll up and down a long list of fittings when the network is defined. The dialog box can be accessed via Library Fittings. The complete set of user-defined fittings which need to be input is shown in table 3.2. Fitting name K-factor PBEND 0.20 D-IN 3.20 D-TEE 0.90 DBEND 0.27 FANIO 2.00 GRILL 5.00 P-IN 0.95 P-TEE 0.48 SEP BAG 3.50 HEPA 3.00 Table 3.2: Fittings to be entered for the ventilation system, using the dialog box shown in in figure 3.3. Figure 3.3: The fittings dialog box, from which the data in table 3.2 may be entered. 4 SSL/TM/0001/01 - c 2006 Sunrise Systems Limited

5 Fan Curve The fan plays a crucial role in the performance of the ventilation system. PIPENET Vision fits a quadratic function to the data points input for the fan performance curve. This is done by a program called the Pump/Fan module. In order to invoke this program we use the command Library Pumps. The fan data for this problem is shown in table 3.3, and is entered into the dialog box shown in figure 3.4. This data will be saved in a library when OK is selected. Flow rate Pressure generated (m 3 /s) (inch water Gauge) Table 3.3: Fan data for example 9, which is entered into the dialog box shown in figure 3.4. Figure 3.4: The pumps - coefficients unknown dialog box, from which the fan data in table 3.3 may be entered. 5 SSL/TM/0001/01 - c 2006 Sunrise Systems Limited

6 The Network To input the network, choose the orthogonal grid option. Then draw the network shown in figure 3.5. The attributes and fittings for the network items are given in table 3.4. Figure 3.5: Network for example 9. Pipe Diameter Length Elevation Fittings label (mm) (m) (m) P-IN P-IN P-TEE P-IN P-TEE P-IN P-TEE P-IN P-IN BAG PBEND HEPA FANIO FANIO (a) Data for circular ducts (pipes) within the network. Duct Height Width Length Elevation Fittings Label (mm) (mm) (m) (m) D-TEE, DBENDx GRILL D-TEE GRILL GRILL D-TEE x 2, DBEND x 2 (b) Data for rectangular ducts within the network. Table 3.4: Data used in the network. 6 SSL/TM/0001/01 - c 2006 Sunrise Systems Limited

7 Note the following: 1. Items 1 to 17 are circular ducts and are input as pipes. 2. Items 18 to 24 are rectangular ducts and are input as ducts. In PIPENET Vision, ducts and pipes are different items and should be input by using different items from the palette. Circular ducts The attributes for a pipe are input by pointing the cursor at the pipe, and data can be entered in the properties window. The property window can be accessed by going to View Properties. An example of a dialog box which has been completed is shown in figure 3.6. Figure 3.6: Properties for pipe labelled 1. This data can be copied and pasted on to pipes 2, 4, 6, 9 and 11. This is done by first pointing the cursor to the source, right clicking on it, clicking on copy, then pointing the cursor to the target, right clicking on it and clicking on paste. 7 SSL/TM/0001/01 - c 2006 Sunrise Systems Limited

8 Rectangular Ducts Figure 3.7 shows the properties window for a duct (in this case, duct labelled 18). Note that the properties window for a duct is slightly different to that of a pipe (figure 3.6), where the width and height of the duct are specified, as opposed to the pipe radius. Figure 3.7: Properties for duct labelled 18. Specifications All the inputs and the single output are assumed to be at 0 in H 2 O G. They can be input by giving the value for one input node, then copying and pasting it on to the others. The specification for node 1 is shown in figure 3.8(a). Fan Characteristics The only other item we need to input is the fan type. This has already been set up and stored in the library. It is simply a matter of selecting the fan from the library. The properties window for our fan is shown in figure 3.8(b). 8 SSL/TM/0001/01 - c 2006 Sunrise Systems Limited

9 (a) Properties window for node 1. (b) Properties window for the fan. Figure 3.8: Calculation and results The data for the problem has now been input completely and so we can proceed to perform a calculation. It is advisable to save all data and perform a check before a calculation is performed. A calculation is performed by using the command Calculation Calculate. The results can also be seen through the browser or Word. If it is examined in Word, all the facilities of Word would be available, including cut and paste etc. The above options can be reached by the command Calc Browser, which leads to the dialog box shown in figure 3.9. Results can be directly displayed in the schematic or in the Properties window. For a detailed excel format output, use the Data window. There is a facility to copy the data from the Data window to an excel spreadsheet (copy/paste command). The Data Window option is ideal for fine tuning the design by looking at the results, making changes to the system and calculating again. Results for the pipes and the ducts are shown in figures 3.10 and Figure 3.9: Output options. 9 SSL/TM/0001/01 - c 2006 Sunrise Systems Limited

10 Figure 3.10: Results for pipes. Figure 3.11: Results for ducts. 10 SSL/TM/0001/01 - c 2006 Sunrise Systems Limited

11 Modelling a Leak Once the initial network has been input, it is easy to perform many calculations to study the various types of failure that can occur in the system. One of these calculations might be to predict what would happen if there was air in-leak due to a perforation. Let us suppose there was a small perforation 20 mm in diameter, exactly half way along duct 20 (see figure 3.12). Let us also suppose that the wall thickness of the duct material is 2 mm. PIPENET Vision can be used to model this by using the following steps: 1. Create an additional node half way along duct Attach a pipe 20 mm diameter, 2 mm (0.002 m) long to this node. 3. Set the pressure at the free end of the new pipe to 0 in. H 2 O G. Now perform a calculation. It can be seen that this only makes a minor difference to the extract flow rates. Figure 3.12: Network, now with a new pipe to represent a leak within the system. Obviously the perforation size can be changed and more calculations performed. 11 SSL/TM/0001/01 - c 2006 Sunrise Systems Limited

12 3.3 Example 2: Balancing a Ventilation System Objectives The objectives of this example are: Select a suitable fan to drive the system. To find ways of ensuring that the pressure remains negative within the compartments. We must also ensure that the direction of flow is from less contaminated areas to more contaminated areas. We also consider what happens if there are doors in between the compartments, which could potentially leak. Finally, we consider the case in which one or more doors are left open Initialisation of data The Network An overall arrangement of the system is shown in figure The compartments are divided into two Figure 3.13: Network arrangement for example 10. because we will introduce interconnecting doors between the nodes in other simulations later. 12 SSL/TM/0001/01 - c 2006 Sunrise Systems Limited

13 Ancillary Data The medium is air at 20 C (use the ideal gas approximation). The units to be used are Pa (G), m 3 /s, m (diameter), m (length). The other units can be chosen as desired. Duct attribures For the sake of simplicity, no fittings are considered. The network data also has a high degree of symmetry so that the copy and paste of attribute data can be used to the full. The data for the ducts is shown in table 3.5. Pipe Width Height Length Elevation Roughness Label (m) (m) (m) (m) (m) Fan Selection Table 3.5: Duct attributes for example 10. In order to select the fans, the first calculation is done without the fans and then the fans will be input. Assume that there are fans at the input and the output. The pressure that needs to be generated at the input is 20 Pa and, at the output, -20 Pa is needed. The pressure specifications are shown in figures 3.14(a) and 3.14(b). (a) Specifications for node 1. (b) Specifications for node 13. Figure 3.14: Input and output nodes for example 10. The fan curves can be calculated by doing a calculation with these initial pressure specifications. PIPENET Vision calculations show that the fans need to generate a flow rate of around 22.3 m 3 /s. Therefore, select a fan with the characteristics shown in table 3.6, bearing in mind that the fans need to be slightly bigger than the minimum required performance. The dialog box for the fan curve is shown in figure SSL/TM/0001/01 - c 2006 Sunrise Systems Limited

14 Flow rate Pressure (m 3 /s) (Pa) Table 3.6: Fan data for generating the fan curve. Figure 3.15: Generated fan curve, using the data from table SSL/TM/0001/01 - c 2006 Sunrise Systems Limited

15 Modelling the System with Fans In order to achieve this, we remove the old specifications (on nodes 1 and 13), add the fans, select the fans from the library (by using the properties dialog box) and set the input and output nodes to 0 Pa. The new input and output nodes are 14 and 15 respectively. The calculation yields the results shown in figure Figure 3.16: Results for the initial calculation, including fans. Balancing the System We note that the flow rate in the middle compartment is a little high and the pressure in the third compartment is positive. (Note that ducts 10 and 13 are slightly shorter than ducts 2 and 5). Our objective now is to find ways of rectifying this. In other words, we wish to reduce the flow rate through the second compartment and reduce the pressure in the third compartment. A suggested solution is to place dampers which are set to drop 6 Pa at 6 m 3 /s. The dialog box in figure 3.17 shows how to do this. The dampers are placed on ducts 6 and 10. It can be seen that the pressure is negative in the sensitive parts of the system and that the flow is more balanced. So, the dampers can be left at the above setting. The results of inlcuding these dampers are shown in figure SSL/TM/0001/01 - c 2006 Sunrise Systems Limited

16 Figure 3.17: Fittings dialog box, from where the parameters for the dampers can be set. Figure 3.18: Results for the network including the fittings on ducts 6 and SSL/TM/0001/01 - c 2006 Sunrise Systems Limited

17 Modelling a Leaky Door It is of interest to evaluate the effect of a leaky door between the compartments. We assume that the door is 2m x 2m in size and has a gap of 2mm around three sides. The door is 50mm thick. We can model this by placing a duct of 6m x 0.002m and 0.05m long between the compartments. The properties window for such a duct is shown in figure In this particular example, the two ducts used to model the leaky door are labelled 15 and 16. The direction of the arrows represents the direction in which the ducts were input. A positive flow means it is in the direction of the duct, and a negative flow means it is opposite to the direction of the duct. Whether we accept this or not depends on which of the compartments are more contaminated. If the direction of flow is not acceptable then the dampers may have to be reset. With a relatively small flow such as this, what is of most concern is the direction of flow. The results are shown in figure Figure 3.19: Properties window for one of the 2 ducts used to model a leaky door. Figure 3.20: Results with the inclusion of a leaky door. 17 SSL/TM/0001/01 - c 2006 Sunrise Systems Limited

18 Modelling Open Doors Finally, it is of interest to see what happens if the doors are left completely open. In order to do this, we simply connect the compartments with 6 m x 6 m ducts of length 0.3 m (assuming that the walls are 300 mm thick). The properties window for such a duct is shown in figure We note from the results shown in figure 3.22 that the pressures equalise between the compartments. However, the pressures still remain negative and are acceptable. The direction of flow between the compartments may or may not be acceptable depending on which compartments are more contaminated. Figure 3.21: Properties window for one of the 2 ducts used to model an open door. Figure 3.22: Results with the inclusion of an open door. 18 SSL/TM/0001/01 - c 2006 Sunrise Systems Limited