Laser-Based Sensors and Algorithms for GCS MVA

For Geologic Carbon Sequestration (GCS) monitoring,
verification, and accounting (MVA), we are developing and evaluating a suite of
cost effective reliable, robust, CO₂ sensors and signal processing
algorithms based on Tunable Diode Laser Absorption Spectroscopy (TDLAS). GCS
pumps CO₂ into  deep brine-filled aquifers under impermeable cap rock
formations. The biggest risk associated with GCS is the potential for
inadvertent CO₂ release from fractures or faults in the cap rock,
injection wells, and transportation infrastructure (e.g. pipelines) with resultant
environmental, health, and safety consequences.  Widespread GCS adoption thus requires
sensors that monitor and verify storage integrity at all depths, from the storage reservoir to the near-surface, and
report if storage properties do not behave as predicted, thus enabling rapid
mitigation.  

To
this end, we are developing gas and liquid sensors that can be permanently installed to continuously
monitor the local CO₂ condition directly within and around the
reservoir.  The sensor technology is based on our well-established Tunable
Diode Laser Absorption Spectroscopy (TDLAS) gas measurement products.  TDLAS
offers flexibility in optical sensor configuration choices and covers a broad
range of gaseous analytes within complex mixtures. For different MVA roles,
differing and complementary sensors are called for, as illustrated in Figure 1. 
They include: 1) long open-path sensors for fast notification and
quantification of containment failures, such as at pipelines or injection
facilities; 2) networked point sensors for continuous and autonomous
measurement with the intent of providing long-term CO₂ mapping and
changes over time of concentrations and fluxes at sequestration site surfaces
and in shallow sub-surface boreholes; and 3) robust fiber-coupled TDLAS sensors
for the detection of supercritical CO₂ downhole in and around the
reservoir CO₂ plume for monitoring the progress and paths of
migrating CO₂. 

Figure 1.  MVA
CO₂ monitoring and accounting sensor suite, including open-path CO₂
gas sensors (red), shallow in-ground CO₂ gas point sensors (blue),
and well-depth liquid CO₂ sensors (green).

Prototype long-path TDLAS sensors in two
configurations were constructed and successfully tested in 2013 at the Illinois
Basin Decatur Project (IBDP) field site: 1) a hand-held, mobile instrument for
standoff leak inspection of CO₂ pipeline infrastructure; and 2) a
solar-powered wireless Open-Path Sensor (OPS) system for continuously monitoring
CO₂ integrated along a 100-m path.  The OPS operated for 12 months at the IBDP field site
with limited maintenance.  It readily detected abnormal CO₂ emissions
near the wellhead related to routine injection operations. These emissions are characterized by rapidly
fluctuating CO₂ concentration signatures; leak detection algorithms
identify these signatures in the face of relatively slow variations in ambient
CO₂ arising from biogenic, anthropogenic, and atmospheric mixing
processes. However, it is likely that CO₂ seepage from GCS
reservoirs will be of slower time scales over areas much smaller than the
reservoir, and thus more difficult to distinguish from natural CO₂ spatio-temporal
variability. Consequently, the monitoring technologies must detect and locate
CO₂ leakage signals of potentially very small magnitude relative to
background CO₂ variations occurring on diurnal to annual time
scales, and occurring over relatively small areas relative to the total area of
monitoring. Collecting data over a distributed spatial area and processing it
over long time periods will be needed.  To this end, we are applying self‑training
algorithms, employing advanced and innovative statistical analyses that process
the data provided by these tools to recognize long-term abnormal signatures
indicating leakage, capitalizing upon the sensitivity, continuous operation,
and affordable distribution of the TDLAS sensors.

             Our all-optical downhole fluids sensor (DFS) concept deploys sensor heads for measuring CO₂ in gaseous or liquid/supercritical phases at reservoir
depth. The sensor head design will be sufficiently small and simple to be
integrated into wireline platforms, or deposited sacrificially on the outside
of well-bore casings, or deployed in  micro (~1? scale) boreholes.  The
envisioned sensor system would have sensor heads deployed in multiple wells
connected to a distributed measurement and monitoring network that spans the
entire sequestration basin.  It fulfills two
distinct roles: 1) monitoring the extent or progress of the injected CO₂
plume in the reservoir; 2) detecting leakage past the cap rock, or, e.g., along
a suspected potential leakage path. These measurements will improve the
quantitative accuracy of reservoir fluid dynamics models, helping to specify
parameters such as plume spread rate, geologic heterogeneity factors, and total
saturation. The second mode represents a critical continuous monitor (alarm)
for protection of neighboring property and aquifers, thus protecting both the
public and the GCS operator.