# Guidance on Estimating VOC Emissions from Cooling Towers found at Petroleum Refineries and Petrochemical Plants

## 1. Background

This guide deals with the estimates of VOC emissions from wet cooling towers in facilities which use heat exchangers to cool hydrocarbons or VOCs in some of their processes. These processes are primarily associated with petroleum refineries and chemical plants.

## 2. Methodologies in Estimating the VOC Emissions

Two principal methods can be used to determine the VOC emissions from cooling towers where the cooling water is used in heat exchangers that come into contact with VOCs such as in petroleum refineries or petrochemical plants:

- Measuring the VOC concentrations in cooling water and obtaining the water flow rate data
- Using the AP-42 emission factors

### 2.1 Measurement Method

The actual measurement method is preferred over the emission factor approach when the data are available. Typically, the measurement method to estimate the emissions involves using a mass balance approach based on two samples taken from the cooling water: one sample before entering and one sample after leaving the cooling tower . The change in the measured concentrations of VOC in the cooling water is then multiplied by the cooling water flow rate to estimate the quantities of VOC emitted into the atmosphere since the last measurement.

The following formula can be used to estimate the VOC emissions from the cooling tower for the time between each measurement:

VOC[t] = (C_{VOC,In}– C_{VOC,Out}) [ppmw] x D_{water} [t/m^{3}] x Water Flow Rate [m^{3}/hr] x H [hr]

Where:

C_{VOC,In} = Concentration of VOC in the cooling water prior to entering the cooling tower

C_{VOC,Out} = Concentration of VOC in the cooling water after leaving the cooling tower

D_{water} = Density of water at 1 kg/L or 10^{6} g/m^{3} or 1 t/m^{3}

H= Number of hours of operation of the cooling tower since the last measurement

If the monitoring is performed more than once during the year, then the VOC emissions calculated for each of the monitoring event shall be summed together to calculate emissions for the year.

The recommended method to measure VOCs in the cooling water is outlined as follows:

- Texas Commission on Environmental Quality (Sampling Procedures Manual): http://www.tceq.state.tx.us/assets/public/implementation/air/sip/sipdocs/2002-12-HGB/02046sipapp_ado.pdf

While the above method is the preferred approach, other methods are also applicable. It is up to the reporter to use a method that is comprehensive, defensible, and applicable for the measurement of VOCs in petroleum refineries and petrochemical plants that may be found in water from cooling towers.

### 2.2 Emission Factor Approach

Emissions | VOC Emission Factors (kg/10^{6} L cooling water) | Applicable Control Technology |
---|---|---|

Uncontrolled | 0.7 | - |

Controlled | 0.08 | Minimization of hydrocarbon leaks into cooling water system; monitoring of cooling water for hydrocarbons |

According to AP-42, use of the controlled emission factor applies to refineries which minimize hydrocarbon leaks to cooling water by ensuring that the water pressure in heat exchangers is at least 35 kilopascals greater than the VOC pressure in heat exchangers or the refineries monitor for hydrocarbons in water.

If there is no regular program to monitor VOC emissions or the means to ensure low emissions, then the uncontrolled emission factor of 0.7 kg/10^{6} L is considered appropriate to use in determining the VOCemissions.

Based on the emission factor method, the following calculation formula can be used for the VOC emission estimates from the cooling tower on an annual basis:

VOC[t/y] = EF_{Unc} [kg/10^{6} L] x Water Flow Rate [m^{3}/hr] x H [hrs] x 10^{-3} [t/kg])

Where:

EF^{Unc} = Uncontrolled VOC emission factor for cooling tower and is equal to 0.7 kg per million litres of cooling water

H = Number of hours of operation of the cooling tower during the year

## 3. Examples of Estimating the VOC Emissions

Example #1:

Sampling during the monitoring period provides results of VOCconcentration in a refinery cooling water of 0.70 ppmw before entering the cooling tower and 0.48 ppmw after leaving the cooling tower. As no leak repair has been conducted between the sampling events, it is assumed these two measurements are representative for the entire year. The cooling water recirculation rate averages 15000 m^{3} /hr. The annual operating schedule of the refinery is 8400 hours. What are the annual emissions of VOC from the cooling tower?

Using the formula from Section 2.1, the calculation is performed as follows:

VOC= (0.70 –0.48) x 10^{-6} x 1[t/m^{3}] x 15000 [m^{3}/hr] x 8400 [hr/y] = 27.72 tonnes/year

Example #2:

An unmonitored cooling tower with a water recirculation rate of 15000 m^{3} /hr is servicing a heat exchanger for the process operations of a refinery, which is operational for 8400 hours on a yearly basis. Estimate the annual VOC emissions from this cooling tower.

The emission factor method is used since there is no monitoring data available. Using the formula from Section 2.2, the calculation is performed as follows:

VOC = 0.7 [kg/10^{6} l] x 10^{3}[L/m^{3}] x 15000 [m^{3}/hr] x 8400 [hr/yr] x 10^{-3} [t/kg] = 88.2 tonnes/year

## 4. Considerations for the NPRI Reporting

The quantity of VOC emitted from cooling towers should be reported as “fugitive releases”.

The determination of VOC emissions by measurement method involves a mass balance calculation, consequently it would be more appropriate to use the NPRI emission estimate code “C” (mass balance) to represent this estimation method for calculating VOC emissions from the cooling tower.

## References

Emission Estimation Protocol for Petroleum Refineries (Version 2.1.1) – Report done by RTI International for the U.S. EPA. May, 2011.

International Workshop on VOC Fugitive Losses: New Monitors, Emission Losses, and Potential Policy Gaps. U.S. EPA. October, 2006.

National Emission Standards for Hazardous Air Pollutants from Petroleum Refineries; Final Rule. U.S. Federal Register, Vol.74, No.207. October 28, 2009.

- Date modified: