Air Cooling and Drying Process

In this article we explore the essential process of Air / Inert Gas Drying using refrigerated units. This process is crucial onboard for applications such as nitrogen generation, inert gas drying, and control air dehydration. Rather than just describing the equipment, the article emphasizes numerical data and calculations. Accordingly, it illustrates examples useful for operation and maintenance purposes.

Table of Contents


Air / Inert Gas Drying using refrigerated units is a crucial application on ships like Gas Carriers. While these components are also common for other applications on ships, such as instrumentation air dehydration, this article focuses on the Air / Inert Gas Drying process using a nitrogen generating plant as an example.

This process involves drying air through cooling as a preliminary step. We will detail the system layout in the next section and demonstrate how simple process calculations can provide essential insights for operation and maintenance.

If equations and thermodynamics seem challenging, there’s no need for concern. Understanding the calculations discussed here mainly requires grasping a few important concepts. Moreover, these calculations can be performed automatically.

To better follow the article, please open the file Refrigerated Drying Process Calc.xlsx. For the best display in its original format, we recommend downloading the file.


The diagram above illustrates the layout of the plant, which consists of three distinct stages:

1)  Air Compression

To distribute the generated nitrogen throughout the onboard delivery network, the system requires pressurization. Therefore, compressing the air is necessary, which also reduces buffering and storage space requirements. Moreover, the nitrogen network must maintain a specified working pressure. In our example, we assume a design pressure of 7 bar gauge for the supply air and nitrogen piping. Minor pressure drops in the system are disregarded as they are insignificant in our study.

2)  Refrigerated drying

We often use the dew point value to quantify the moisture content of the air. Basic instrumentation air applications typically require a dew point of approximately -20°C. However, our nitrogen generating plant requires a lower dew point to ensure minimal water content in the cargo tanks. The refrigerated drying unit in our example cannot achieve dew points between -40°C and -70°C as required, but initially reducing the dew point to around -20°C is crucial. That way, we ensure that subsequent drying processes remain effective.


3)  N2 Generation

In our example, the third stage is the N2 generation system, which primarily consists of two PSA vertical vessels. These vessels remove oxygen and other unwanted gases from the incoming air. PSA, or Pressure Swing Adsorption, is one of the two main N2 generation technologies. These systems use a Carbon Molecular Sieve (CMS) to capture oxygen and unwanted gases while allowing nitrogen through. Another common nitrogen generation method is the membrane system, but this article will not cover it.

One PSA vessel operates in “capture mode” for several hours, while the other operates in reverse flow, releasing captured oxygen, CO2, and other gases back to the atmosphere. This process regenerates the CMS for a new working cycle. After a set number of hours, the vessels switch from capture mode to regeneration mode, and vice versa.

Are you familiar with above layout?

Readers from the maritime sector are likely familiar with the described layout, often installed using three different skids. Note that the provided schematic is simplified, showing only the main components for clarity. While other designs may implement variations and include additional components, they are likely based on the same principles. The diagram depicts a simplified example of an Air / Inert Gas drying application you may recognize. In this case, the media undergoing drying is air that the system uses to generate dry nitrogen as inert gas, hence the phrase “Air / Inert gas Drying”. Regardless of your familiarity with this specific setup, we welcome your comments and feedback.
Now, let’s delve deeper into the components and functionality of the Air / Inert Gas Drying unit before discussing the benefits of process-related calculations.


Once we understand the role of air / inert gas drying in the overall process, we can focus on the refrigerated drying stage. Examining the diagram reveals the basic components of a refrigeration cycle: compressor, condenser, expansion valve, and evaporator. One might wonder why cooling is necessary. Cooling air causes most of its moisture content to condense into liquid form and drain out. As temperature decreases, air’s capacity to retain water vapor also decreases, leading to excess moisture condensing on cooler surfaces, much like how water condenses on car windshields in winter.

Moisture removal occurs in the evaporator (11). For those unfamiliar, let’s clarify the terminology: it’s called an evaporator because refrigerant evaporates inside the tubes, causing a sudden temperature drop. It’s not the water that evaporates! Instead, moisture in the air passing over the evaporator’s exterior cools and condenses. This condensed moisture drips into the evaporator housing and drains out via the automatic drain valve (17). Whether in a car’s air conditioner or a home refrigerator, the evaporator is where cooling occurs.

That heat exchanger there…

One component in the diagram that often raises questions is the air pre-cooler (10). Following the airflow path, before entering the evaporator (11), the air passes through the pre-cooler (10) where it exchanges heat with the colder air previously refrigerated in the evaporator. This arrangement may seem counterintuitive, but there’s a rationale behind it. To grasp the concept, consider what would happen without a pre-cooler.

Assuming no pre-cooling:

In this scenario, the air enters the evaporator at 54 °C, and exits at 5 °C. The temperature at the evaporator should be above 0 °C, otherwise moisture would freeze on the heat exchanging tubes. As a result, the ice build-up would block the airflow across the evaporator. Manufacturers typically include a safety margin of approximately 3-5 °C to prevent such issues. While optimizing the drying process by setting a lower evaporating temperature, such as 1 or 2 °C, is feasible, careful monitoring is essential due to potential instrumentation errors.

To summarize, the evaporator capacity must be 37 kW to reduce the temperature from 54 °C to 5 °C. (Refer to the “Heat Exchange” sheet in the excel file Refrigerated Drying Process Calc.xlsx.)

Once the air exits the evaporator at 5 °C, re-heating to a specific temperature is necessary for several reasons. Despite the advantages of cold temperatures in pressure swing adsorption, manufacturers may recommend a specific operating temperature for IG. Moreover, excessively cold temperatures may not be suitable for inert gas applications. Warmer inert gas temperatures allow for greater moisture absorption before purging from the tank. In our example, we assume a design temperature of 30 °C for the air before entering the N2 generation stage. Therefore, we need to provide 12 kW of heat to raise the air temperature from 5 °C to 30 °C.

Now, let’s examine how the pre-cooler contributes to better efficiency:

Put simply, the pre-cooler utilizes previously cooled air to lower the temperature of incoming air from 54 °C to 42 °C. This equates to 12 kW of free cooling, as calculated in the spreadsheet. Consequently, the only energy required is 37 – 12 = 25 kW to cool the flowing air mass from 42 °C to 5 °C. No energy is needed to reheat the cold air from 5 °C to 30 °C because the same effect was achieved at no extra cost while pre-cooling the hot incoming gas.

Importance of the drain valves

It is crucial to keep the automatic drain valves in good operational condition. If they become clogged, stuck, or fail to drain condensate water for any reason, the undrained moisture may reach and potentially damage the PSA units. Exercise No. 1 below will demonstrate how we can check whether the drain valves are effectively performing their task.


Below, we present two calculation exercises (Practice 1 and Practice 2) for common onboard situations.

Practice 1

The first exercise involves calculating the amount of water expected to drain before air enters the PSA units.

Why is this practice beneficial? By comparing expected and actual amounts of drained water, we verify proper function of Stages 1 and 2. If discrepancies occur, the issue likely lies in the PSA units.

Matching the drained water to expected amounts also confirms correct airflow rates. For example, if the first drain matches expectations but the second does not, it suggests an issue with the pre-cooler or its drain valve. Calculating the expected amount of drained water offers clear advantages. Measuring the actual values is straightforward using a bucket and a watch.

Practice 2

In this exercise, we calculate the dew point of air before it enters the PSA units. This is useful if a dew point meter is unavailable or for verifying its accuracy.

Before starting calculations, we need to note the relevant initial conditions and design parameters shown in the table below. We need the RH value of ambient air; the simplest way to obtain this is by using a hygrometer. If a hygrometer is unavailable, we can estimate RH using a dry bulb thermometer and a wet bulb thermometer. For brevity, we assume a hygrometer is accessible.

Initial data include:


Air Inert Gas Drying Process Initial Data


Problem definition

Calculate the mass of water in kg/h (or kg/s) expected to condense and drain from the system.


We’ll start with the initial ambient conditions and three additional points where air moisture content varies:

  1. AMBIENT STAGE: Represents the initial air conditions at compressor suction. This stage is crucial as it establishes values needed for subsequent calculations.
  2. DRAIN VALVE 9: Water condenses at this point due to increased air pressure and cooling. Note: The exact location of the drain valves might vary, but the expected amount of condensate should not.
  3. DRAIN VALVE 18: Located immediately after pre-cooler 10. Condensation occurs here due to the decrease in air temperature, which reduces the amount of moisture the air can retain.
  4. DRAIN VALVE 17. An additional amount of water drains at this point. Again, this condensation occurs due to the decrease in air temperature.

Comments and clarifications

An explanation of the calculation procedure follows below. First, we recommend opening the relevant Excel file in a separate window by clicking here. If your browser does not display all the details correctly, you can download the file instead to view its original version.

Please note that the spreadsheet calculates values using REFPROP software. If you do not have the software installed on your computer, the formulas will show an error condition. Therefore, we have replaced the formulas with calculated values shown in blue. Otherwise, If you have REFPROP integrated into your Excel application, feel free to Contact Us, and I will share the formulas with you. Also, note that we do not have any commercial relationship with REFPROP. We are simply users of the software, much like many others in the industry and academic circles.

The spreadsheet may seem daunting to those unfamiliar with the subject matter. However, once you grasp the logic, the calculation process will flow naturally. You’ll quickly be able to replicate the procedure for similar problems. As they say, it’s more complex to explain than to actually perform!


To understand the logic behind the spreadsheet, let’s revisit the process used to prepare it. We start by setting up a small table (framed in blue) with the gas components of dry air, their molar fractions, and their molar masses. The cells in light brown contain fixed values sourced from chemistry literature. In contrast, the green cells contain values that depend on the moisture content of the air. Consequently, the spreadsheet calculates these values for our specific conditions. The key values that initiate the calculation process are also framed in blue. Their inter-relationship is as follows:

  • Using the values of relative humidity (RH) and the saturation pressure of water in air (PSA), we determine the partial pressure of water in ambient air (PPW) through a simple proportionality: PPW = PSA * RH. Typically, tabulated data in chemical literature provides the saturation pressure for water vapor. However, we use REFPROP software for convenience in our calculations.
  • Dividing the mentioned partial pressure by atmospheric pressure gives the molar fraction of water (MFW) in ambient air. The spreadsheet then incorporates this value into the small table alongside the previously mentioned air component values. Simultaneously, the spreadsheet calculates the molar fraction of dry air as MFD = 1 – MFW.
  • After determining MFW and MFD, the spreadsheet automatically updates the remaining values highlighted in green in the framed table. This process accurately defines the air mixture for moist air. With this information, REFPROP calculates the density of the moist ambient air.


We continue to discuss the calculation of mass flow rates for ambient air drawn by the compressor:

  • In point A) above, we determined the density of ambient (wet) air. In our example, the compressor has a volume flow rate of 1500 m³/h. Multiplying the volume by the density gives the mass flow rate (kg/h).
  • Next, we convert the mass flow rate to mol flow rate by dividing it by the molecular mass of the wet air, which we calculated previously.
  • Applying the molar fraction of dry air in ambient air to the calculated value above gives the molar flow rate of dry air. Converting this value to the mass flow rate of dry air is straightforward, as we already know the molecular mass of dry air. We can then determine the mass flow rate of water by subtracting the mass flow rate of dry air from that of wet air. Calculating the concentration of water per kg of dry air is also straightforward.

The mass flow rate of dry air introduced above plays a crucial role. Initially, one might consider basing calculations on the mass flow rate of the actual circulating wet air. However, attempting such an approach would reveal that the mass of wet air varies each time water drains, complicating calculations. In contrast, the flow rate of dry air remains constant regardless of how frequently water drains. Therefore, using this parameter as a reference is highly convenient.


At each drain point where water condenses, the water vapor is in a saturated condition. The calculation method begins with the first drain point, located at the buffer vessel. Previously, we discussed how to determine the saturation pressure of water. Since the moist air is saturated within the buffer vessel, its saturation pressure equals the partial pressure of water. Dividing the partial pressure of water by the total pressure of the wet air gives us the molar fraction of water in the buffer vessel.

Subsequently, the spreadsheet can calculate the green cells in the light brown block similarly to the method explained earlier. This process is now simpler due to the availability of the mass flow rate value of dry air, eliminating the need to use the air density value. As discussed earlier, the water content of air is expressed in kilograms of water per kilogram of dry air.

Finally, the amount of water extracted at the first drain point is determined by the difference between the initial amount of water and the remaining amount of water in the air. The subsequent stages (drain points) follow a similar calculation process.


After completing the described procedure for each drain point, we obtain the following amounts of water drained:

  • Buffer Vessel Drain valve: 16.38 kg/h
  • Pre-cooler Drain valve: 9.19 kg/h
  • Refrigerated Drying Unit Drain valve: 9.72 kg/h

Total amount of water drained: 35.29 kg/h

Remaining water content: 1.16 kg/h


Problem definition

Calculate the dew point of the air before the PSA unit.

Comments and clarifications:

This exercise is highly beneficial when your dew point meter is not operational or when you need to verify its readings. Another advantage of calculating the dew point temperature for specific conditions is that it eliminates the need to run the system under those conditions or be present at the installation.

It’s important to note that we’re discussing the atmospheric pressure dew point in this example, not the pressure dew point. The pressure dew point pertains to the temperature at which pressurized air becomes saturated, while the atmospheric pressure dew point assumes air at atmospheric conditions and refers to the temperature at which it becomes saturated.

From Practice 1, we determined that the air from the refrigerated dryer is at 7 barg (0.8 MPa) and saturated at 5°C, making the pressure dew point for those conditions 5°C. Now, let’s calculate the dew point at atmospheric pressure, which is more practical since the inert gas produced from the air will be used at atmospheric pressure.


For this exercise, we will utilize values calculated in the previous practice. Please refer to the ‘Dewpoint’ spreadsheet in the Excel file ‘Refrigerated Drying Process Calc,’ which you likely downloaded for Practice 1. The initial conditions in the yellow cells are replicated here for convenience, along with relevant results calculated for the last drain point in Practice 1. Yellow highlights indicate cells with known values.

To calculate the dew point temperature, we will use the formula shown in the light green box. This formula relies on two constant values and a variable, alpha, which depends on relative humidity and temperature. Given that the air temperature out of the refrigerated dryer is 5°C, we will use this temperature in the formula. Therefore, the only remaining task is to calculate the relative humidity.

The strategy is straightforward: we define relative humidity as the ratio of the partial pressure of water at a specific temperature to the saturation pressure of water at that temperature. The partial pressure of water is calculated as the molar fraction of water in air (a known value) multiplied by the atmospheric pressure (another known value).

Thus, we find that RH = 12.6% = 0.126.


Now, we simply enter the relevant values into the formula. The dew point formula gives a result of -21.5 ºC.

Interesting Facts

We calculated the dew point using data from the outlet of the refrigerated dryer. Alternatively, we could have used different conditions, like a temperature of 30°C (before the PSA units). The resulting RH and dew point would be the same. The dew point remains unchanged when air warms; it depends solely on its water content, which doesn’t change when warmed. However, the dew point does change when air cools and water condenses and is removed.
Another useful fact is that the dew point can be calculated without a formula, using a psychrometric chart. We may explore psychrometric charts in a future article, though we currently lack one due to copyright concerns.
While psychrometric charts are valuable, calculating dew point on a spreadsheet offers clear, accurate results without the potential for error from chart interpretation and interpolation.


Our intention has been to provide insight into air / inert gas drying systems commonly used onboard ships. We reviewed a type of marine installation that is quite common, especially on gas carriers. Such a system consists of a compression stage, a cooling stage with a refrigerated drying unit, and a nitrogen generating unit. The system was briefly described, and we explained the reasons for installing a pre-cooler.

We took a unique approach to discussing marine equipment, highlighting how thermodynamic concepts and theoretical calculations can be applied to operations and maintenance activities. Importantly, we demonstrated how seemingly irrelevant data, such as the amount of drained water, can provide useful information.

We welcome your thoughts and invite you to comment.