Fans in a series, parallels. As an approximation it may be said that when fans are connected together in series then, at any give volumetric flowrate, each fan adds its corresponding fan total pressure to the combined output with its corresponding power. In actual practice there is a slight loss in pressure in the connections between the stages. In more exact work it should be noted that the total pressure of the combination is equal to the sum of the fan total pressures of the individual units minus the losses in the interconnecting duct.

# Fans in a series, parallels

Thus the fan static pressure of the combination is equal to the total pressure of the first stage plus the static pressure of the second stage there being only one velocity pressure lost at the final outlet. With high pressures compression becomes important. The second stage will receive its air at a density increased by the pressure of the first. Due to this increased density its pressure development will be correspondingly greater, together with its absorbed power. For normal commercial requirements, series operation is in use mainly for air supply to furnaces, which require a relatively high pressure at a small air flow. Two stages meet most needs, but a larger number of stages may be used for applications such as industrial vacuum cleaning, pneumatic conveying etc.

The transition point will vary from one system to another according to the amount of laminar flow present due to low velocities at filters etc. Only pneumatic conveying plant, dust exhaust and high velocity air conditioning are likely to have flows which are fully turbulent. These effects should be recognized especially when speed control is included.

## Fans in parallel

For a given system total pressure the volume delivered by the combination is the sum of the individual units at the same fan static pressure. This is only strictly true where the two fans are connected to a chamber. If the fans blow directly into a common duct then neglecting losses, the volume delivered by the combination for a given total pressure is the sum of the volumes delivered by the individual fans at the same fan total pressure. Multivane forward curved bladed fans are not usually suitable for parallel operation due to the shape of the fan curves. The stall of low volumetric flowrates means that there may be as many as three flowrates, where the fan pressure is the same.

Because of the pronounced peak in the pressure/volume curve, where there is any possibility of large and rapid fluctuation in system resistance, a forward curved fan selected at any pressure Q above the dotted line (see Figure 5.7) can be unstable. If, for any reason, the flow drops the point of operation can move from something normally around B to C where the fan head is slightly less.

The change in volume may have been small and the system back pressure will have stayed almost unaltered. Thus the system pressure will be in excess of the fan pressure causing the flow to decrease rapidly back to A. Since the back pressure is still above the shut-off pressure a reversal of flow can occur.

The system is then at a standstill and the system pressure (which we assume is oc Q2) now drops below the shut-off pressure. Volume flow increases and the operating point moves up the curve past the equilibrium point. It then comes back and may tend to overshoot, thus repeating the cycle. Such behaviour is accentuated at higher pressures, on long duct runs or when the fan discharges into a chamber of large dimensions. The instability is often not found during normal fan performance tests as these conditions do not then exist.

It will be seen that the practice of selecting over-large fans for a system to reduce the outlet velocity can be extremely dangerous. It may even lead to operating points to the left of the peak pressure B which should be avoided under all circumstances. It is usually necessary to operate identical fans together to ensure that each does an equal share of the work.

## Incorrect rotation

This is common particularly for fans with the impeller mounted directly on the motor shaft extension. In this arrangement, with ducts fitted on inlet and discharge of fan, it is not easy to see any rotating part. Observation has to be made on the shaft as seen down the gap between the motor and the fan. This mistake can arise when the erector leaves the job before it is wired. Many people think that if a fan runs in the wrong direction it will “blow from where it should suck”, which is of course not true. It is important to note that in some installations the reduced flow due to incorrect rotation is not obvious to the customer.

Hence if the job is wrong and not checked he may not complain but in time will be dissatisfied with the work. Examples from experience will illustrate this. In a sawdust collecting plant a backplated paddle fan handled 1.65 m3/s with incorrect rotation and actually worked in a poor manner. When corrected the flowrate was 2.41 m3/s. Other sawdust collecting plants have given similar results.

A paddle bladed centrifugal fan was installed for handling exhaust from paint spraying booths with a textile bag filter on the discharge. It was put into operation, with another similar plant, with incorrect rotation. They worked this way for some time until a visit was made and the fault noted. The volumetric flowrate was 2.029 m3/s as compared with 3.303 m3/s when corrected.

The only means of checking by the customer was the feel of the air entering the booths. It was designed for a face velocity of 0.825 m/s but in the wrong fan rotation was about 0.5 m/s. As 0.5 m/s is common for cut-price work, it is easy to see that a customer might never complain, although not satisfied. Narrow cast iron centrifugal fans are liable to this mistake. A 225 mm fan on a small job handled 0.035 to 0.038 m3/s in the wrong rotation and 0.069 m3/s when corrected. A cast iron fan with forward curved bladed impeller handled 81% of specified flow with power about the same either way One case is known of a cast iron fan which had been running in the wrong direction for seven years before it was noticed! On forward curved multivane fans the wrong rotation is obvious as the flow is so much reduced and cannot fail to be noticed. The same applies to wide backward bladed fans.