Noisy Engines - Building Ventilation Without Open Doors
When your operation is controlled by the AEUB Noise Control Directive (ID 94-4) it is often an acoustic requirement that a compressor be run with all doors and windows closed, in order to contain the engine or turbine casing mechanical noise. A typical compressor building wall will provide an Insertion Loss (IL) of approximately 20 to 30 dB (the IL is the difference between the noise at a given receptor with no building and the noise at that receptor with the compressor building in place). However the wall is not the only component of the building and noise will leak out of the building through the paths of least resistance. Louvers, doors (particularly roll-up doors) and windows all provide leakage paths through which noise escapes. Open doors, windows and louvers create major potential noise - containment deficiencies. For example, a building designed with an IL of 25 dB will have its effective IL reduced to 15 dB if only as little as 3% of the total wall surface area is comprised of open doors, windows or louvers.
During the summer operators prefer to leave all doors, windows and louvers open to increase building ventilation in spite of environmental noise requirements. Closing these ventilation openings only makes the typical operation unbearably hot. The only solution, if an operator has an overheating building problem simultaneous with an environmental noise problem, is to force-ventilate the building. These ventilation fans and intake louvers may themselves need to be equipped with duct silencers so the ventilation air may pass through while containing engine noise within the building. It is imperative that the operator keep the doors shut as open doors merely circumvent the installed noise control treatment.
If the location is remote and electric power is not available, then the operator may consider "bleed ventilation" using the idea in the attached sketch. A tap off duct between the fan and the cooler bundle bleeding ½ to 1% of the fan capacity would typically give 5 - 10 air changes per hour of ventilation. The diagram is shown for a forced draft cooler but works equally well with the induced draft design. In the latter case, the tap off would suck air out of the compressor building. This may then allow the operator to keep the building doors and windows closed thereby reducing the noise attributable to sources located in the building to the surrounding environment.
Overheating Engines
Every summer (and sometimes earlier) many operators face the problem of overheating engines. Engine jacket water (EJW) systems supposedly designed for ambient temperatures in excess of 30ºC often appear inadequate as ambients rise into the mid twenties.
Heat transfer problems may relate to airside deficiencies, tubeside deficiencies, fouling or a combination of all three. The chart below illustrates how the overall heat transfer rate (U value) is related to the airside film coefficient, tubeside coefficient and fouling factor (ff). Most EJW cooling systems are design rated with U values in the range 130 - 170 BTU/hr.f2.ºF. However, our field tests of malperforming units often show U values well under 100.
Typical airside problems relate to fan pitch, fan speed and/or fan tip clearances being incorrect however these problems may also relate to excess backpressure due to plugging of bug screens or fins, damaged or bent fins etc. Typical tubeside problems relate to internal fouling of tubes and/or restricted glycol flow rate.
Regardless of the reason for the overall heat transfer deficiency, there is a limit to which these can be remedied on the airside only. Boosting fan air throughput has two advantageous effects. Firstly the air film coefficient is improved and secondly the log mean temperature difference (LMTD) is improved by reducing the hot end temperature difference. Referencing the heat transfer chart illustrates both the advantageous effect and also the limitations of a strictly "airside fix". In the example charted, if it were possible to boost the air flow by 50% (compared to as tested performance) then the air film coefficient would increase 25% and the airside temperature rise would decrease by 33% with the resulting improvement on the hot end approach. This increase of airside film coefficient would typically improve the overall U value by 8 - 10%. All of this has a positive effect on the performance of the cooler, however if the essential problem is dominated by tubeside deficiencies then the law of diminishing return quickly takes over. The remedial approach of increasing airflow may be practically limited.
The tubeside remedial approach may be more pertinent. In the case of EJW coolers, these units are usually designed to keep the glycol/water mix in a turbulent flow regime. Turbulence is dictated by the Reynolds' number which in turn is governed by average fluid velocity, tube diameter and fluid viscosity.
Reduced fluid velocity may be a result of a malperforming EJW pump; plugged tubes or a defective engine thermostat. Any or all of these will negatively effect the EJW Reynolds' number.
The viscosity of a given glycol/water mix is fixed, however the operator can manipulate this ratio to some degree to his advantage during the summer. Most EJW systems are designed for 50/50 glycol/water. This provides deep freeze protection in the winter. However in the summer this glycol strength is not necessary and operator may beneficially alter this to 30/70 glycol/water. This still allows enough freeze protection for the summer months but provides improved heat transfer properties than a 50/50 mix. The danger with overdoing this remedy is that the vapour pressure of the mixture increases at a given EJW temperature.
In summary, we suggest that a proper field test of an overheating engine jacket water cooler is usually a wise move before attempting "quick fixes" such as fan alterations, auxiliary coolers, rebundling, larger EJW pumps, tube accelerator rods, etc. While any one or a combination of these ideas may be the best solution, there are so many variables that the chances of getting it right without a field test are difficult.
