Menu

Tank Emissions: Taking care of environmental business

By Angela Solano and Herbert Loria – VMG Calgary     

Introduction

Governments worldwide are making stronger commitments to monitor and regulate air quality. This places higher demands on the oil and gas industry to report air emissions and meet tighter regulation targets at various stages of operation. In this article, we show you how emission prediction and reporting are seamlessly integrated with your plant model in VMGSim 9.0. “Doing your part” for the environment has never been so easy. 

VMGSim 9.0 provides all the tools necessary to calculate air emissions alongside your design, so these calculations can be updated and analyzed automatically. The newly built-in Tank Emissions feature takes care of estimating emissions from any storage tank in the model. It gives users the ability to compare emissions from alternative tank designs, change ambient conditions, run case studies, add variables to Visio PFD, extend calculations in spreadsheets, generate emissions reports, and more; all inside the VMGSim case. Flash emissions from any separator can be calculated and analyzed as well, with rigorous fluid and oil characterization that maximizes the accuracy of these predictions.

Background

Environmental agencies classify air pollutants in a variety of ways. Hazardous air pollutants (HAPs) are pollutants that have been linked to serious health, environmental or ecological effects. These air toxics include benzene and carbon disulfide. Pollutants like sulfur dioxide and nitrogen oxides are known to cause acid rain. Greenhouse gases like methane and carbon dioxide contribute to climate change. The oil and gas industry is the largest industrial source of emissions of volatile organic compounds (VOCs), which contribute to the formation of smog and are thereby linked to a wide variety of health effects.

Sources of emissions in oil and gas plants include flares, venting from separators, and breathing and working losses from storage tanks. Flash emissions occur when vapor is released from a fluid as a result of a pressure drop, and this vapor is allowed to escape to the atmosphere; for example, during the transfer of a fluid from a separator to a vented storage tank. Breathing losses occur as a result of vapor space breathing in the storage tank, as changes in atmospheric conditions cause the vapor in the tank to expand and contract. Working losses occur during tank filling or emptying operations, as liquid clings to the exposed surface area in the tank and eventually evaporates.

Total emissions from a site have to be estimated and reported to an environmental agency, often before a building or operation permit can be obtained. Emissions are reported in the form of emissions factors, developed by the United States Environmental Protection Agency (EPA) to quantify the pollutants released to the atmosphere as a result of an activity. These values are meant to convey representative long-term averages of emissions for each type of source. The methodology for emissions factor calculation was published by EPA in AP-42, Compilation of Air Pollutant Emission Factors.

Case Study

This case study will demonstrate the basic steps needed to estimate and report flash and tank emissions using VMGSim. It will then touch on how other features of the simulator can be used to analyze, optimize, and report emissions in the model.

We will start by calculating the flash and tank emissions of a crude oil storage unit, followed by the emissions from storage of downstream products.

Part 1: Single Tank Emissions

Fluid Characterization

Emission factors calculations are highly sensitive to stored fluid properties, particularly vapor pressure. These properties are rigorously calculated in VMGSim using equation of state or activity coefficient models. The program offers a variety of thermodynamic models tuned for different industries and applications, and the right selection for the system allows for more accurate estimates of flash and tank emissions.

Proper fluid characterization is essential when modeling oils and heavy hydrocarbons. VMGSim has the capability to characterize oil using either the traditional pseudo-component approach or the new PIONA (n-Paraffin, Iso-paraffin, Olefin, Naphthene, and Aromatic) approach. The latter is the state of the art in its ability to match experimental properties of a fluid even with limited laboratory data.

In this example we use the Advanced Peng-Robinson thermodynamic model for fluid property calculations, and the PIONA approach to characterize the crude oil. An upcoming newsletter will delve into the details of this particular PIONA characterization, and following is a brief preview of the data and results.

We set up an Oil Source unit operation to enter laboratory data for the reservoir: a C7+ analysis, MW, API gravity, and the reservoir saturation pressure. After setting operating conditions for the reservoir stream (temperature, pressure, and flow) we click “Regress Parameters” to match the oil properties.

image001-0008.png

Once the regression is complete, the PIONA oil composition is available, and the Oil Source can be connected to a material stream that we name “Reservoir_Oil”.

image003-0008.png

We may now use a PVT Analysis unit operation to validate the regression. This unit operation analyzes the behavior of the oil by means of the simulation of some PVT experiments, like Differential Liberation or Separators, and we are then able to compare these results to available experimental data. Details on this functionality will be demonstrated in a future newsletter article. The following graph shows the result of a separator gas analysis on the oil in VMGSim compared with experimental data. The match of results was similar at 100, 200, and 300 psig.

image005-0007.png

Characterization that predicts fluid behavior so accurately is fundamental to obtaining better emission estimates.

Flash Emissions

Flash emissions that occur during the transfer of the crude oil to the storage tank can be calculated using a separator. We add a heater in between the reservoir oil stream and the separator to simulate the change in conditions during the fluid transfer.

image007-0008.png

The separator is set to be at atmospheric pressure and the average flash gas temperature, 90F. The vapor port of the separator represents the flow of gases that is vented to the atmosphere, while the liquid port represents the composition of the liquid that is stored in the tank.

image009-0006.png

Flash emissions are the mass flows of the vapor port. We can customize how flash emissions are presented from here, for example adding them to a PFD Datasheet, exporting them to a process calculator, or exporting them to excel. The display units can be set in any of these interfaces to match the reporting units for flash emissions.

Below are examples of how these values can be displayed on a data sheet or used in a process calculator. Emissions are presented in metric tons per year.

image011-0007.png

image013-0008.png

image015-0007.png

Storage Tank Emissions

Emissions that occur during tank breathing or liquid filling/emptying operations can be estimated using the Tank Emissions utility. This tool is accessible through the “Tank Emissions…” button in the separator.

image017-0005.png

The first objective in the example is to estimate the emissions that would be vented from a vertical fixed roof tank holding the crude oil.

An emissions case is created by default when we first open the Tank Emissions tool. We rename this “Emissions_Case” to “Emissions_VFRT”.

Tank design parameters are input in the Main Data tab, including tank type and dimensions, paint details, and vent settings.  The stored liquid is defined as a Crude Oil and its composition is read from the liquid port of the separator, Liq0. The tank is heated, so the liquid bulk temperature is directly specified as 85 F. Ambient data is loaded from the database as we specify that the nearest city to the tank is Houston, Texas. The tank has 80 turnovers per year.

image019-0002.png

Overall emissions results are shown in the Emissions Summary frame on top of the tabs. These are broken down into standing storage losses and working losses and calculated on an annual basis. The Time Basis parameter can be changed to calculate emissions over a monthly time period instead.

Detailed results are presented in the Detailed Emissions tab. Values like vapor space volume, vapor pressure of the stored liquid, and losses per component are available here.

image021-0003.png

All specs and calculations for the emissions case can be exported to an excel template directly from the Report tab.

Alternative Designs

The results estimate 0.14 tpy of propane emissions from this tank. We want to analyze the effect that using a floating roof tank instead of a fixed roof tank would have on this value.

A new emissions case is created with the name “Emissions_EFRT” and the new design parameters are input in the Main Data tab.

image023-0004.png

Initially there is no fitting loss because fitting information has not been added. This can be done in the Fittings tab, where for this example we elect to use the AP-42 recommended set of fittings based on the tank type and dimensions. This list could be customized after deselecting the option to “Use Recommended Set”.

image025-0004.png

There is now a fitting loss. The design specification is complete and results can be compared. The floating roof is predicted to decrease propane emissions over 50% from 0.14 tpy to 0.06 tpy.

Values in the Tank Emissions tool can be used like values in any unit operation. For example, they can be added to a PFD Datasheet or exported to a process calculator as shown below.

image027-0002.png

image029.png

These variables can also be used in case studies to analyze the effect of process or operating conditions on emissions. In this example we do a simple case study to view the effect of manipulating the number of annual turnovers on tank emissions.

image031.png

image033.png

Part 2: Plant-wide Emissions

VMGSim allows users to calculate, analyze, and report emissions from all sources in a site using a single simulation case. Part 2 of this case study will show what this might look like in a basic simulation model, and then a more complex model. 

Emissions from crude oil storage were estimated in part 1; in part 2 we will be estimating emissions from storage of the downstream liquid products. Since products heavier than naphtha are unlikely to have significant emissions of the lighter pollutants in the case, we will focus only on the storage of naphtha and straight-run naphtha.

We can use component splitters to model ideal distillation of the crude oil by boiling points. A tower unit operation could be used to obtain rigorous results, but at this point the goal is to get a rough approximation of the stored liquid composition in each tank.

The flowsheet is expanded as shown below, with the stored liquid stream from the crude oil tank connected to a component splitter.

image035.png

We use the “Assign Splits by Group” option in the component splitter to separate products by boiling point. The bottom splitter sets all components with a normal boiling point higher than 380 F to be heavier products.

image037.png

All other components are sent to the second component splitter to be separated at 200F, into naphtha and lighter products. 

Finally, a separator is used to model the partial condenser at the top of the tower that separates straight-run naphtha from light ends. The condenser is assumed to operate at 100 F and 20 psia.

Product storage tanks are modeled by connecting separators to the naphtha streams. Each separator is specified to operate at atmospheric pressure and flash gas temperature. A cooler is used where required to model changes in fluid conditions during transfer to the tanks.

image039.png

Flash emissions and storage tank emissions can now be calculated for each tank as shown in Part 1 of the case study.

Results can be analyzed and presented in multiple ways, for example shown in PFD data sheets or compiled in a process calculator.

image041.png

image043.png

Simple models like this can provide initial estimates for plant-wide emissions at early stages of a project. After this initial setup the model is free to continuously evolve as more field data and/or design specs become available, and emission estimates will be updated automatically with the model.

It is just as easy to add emission estimates to a complete and rigorous plant model. For example, open the PIONA fluid catalytic cracking plant example in VMGSim. This is accessible through the main menu’s File > Plant Examples.

image045.png

Add a new subflowsheet in the case and set up a storage tank in it for the plant’s gasoline product. Spec the separator at atmospheric pressure and flash gas temperature, and connect the “Gasoline(145)” stream from the main flowsheet to the cooler.

image047.png

Flash emissions from the tank transfer are now calculated and available in the separator’s vapor stream mass flows, and the tank emissions can be estimated using the Tank Emissions tool in the separator.

This simple process can be repeated for any liquid feeds or products in the case that go to storage. The rigorous model is expected to give precise emissions estimates.

Sources 

  1. “National Air Pollutant Emissions”, Environment Canada, 9 May 2014. Web. 24 Jan. 2015.
  2. “Oil and Natural Gas Air Pollution Standards”, US EPA. Web. 22 Jan. 2015.
  3. US EPA, Document AP-42 Chapter 7, Organic Liquid Storage Tanks, Fifth Edition, 2006
  4. McCain, W.D. The Properties of Petroleum Fluids, PennWell Books 1990
To Top