My two half cents:

The low frequency content is critical for accurate source and focal mechanism reconstruction. Remember that it is the least attenuated part of the energy. Yes, it will contain lots of noise but it does improve filtering processes and wavelet features extraction.

The other reason is that even if you low cut your data above the seismometer band you can still use it to model a wavelet in which you assure you have no signal leak or aliasing.

Buona  giornata ed a presto

Bernal

Bernal, thank you for your answer.

it is ok for the first point, even if for the focal mechanism it is important also a good coverage of the focal sphere and at the cost of a broad band station it could be possible to deploy at least a couple of short period stations enhancing the coverage.

Sorry, but I miss your second cent: what do you mean by "...low cut data above the seismometer band..."? Please can you elaborate in more details?

Ciao,

Lorenzo

Dear Lorenzo,

The availability of VBB technology is a power tool for characterization of seismic source and moment tensor inversion across the whole frequency band. The installation of VBB seismometer is very hard and expensive and the site must be selected with low background noise and with a shield from atmospheric effects as variations in pressure and temperature. A good coupling between the sensor and the ground rock (rigid) is essential. In Sicily a VBB station was installed in an abandoned railway tunnel and inside we realized two rooms: one for the VBB seismometer and the other for data acquisition system.

I suppose the goal of your network is to detect and study the local (micro)seismicity with low magnitude (

Fabio, yes in my original question I forgot to specify that we are interested to study microseismicity with Magnitude < 2.

Ciao,

Lorenzo

Lorenzo,

I suppose you know that the government offices are writing the guidelines about seismic monitoring networks around oil/gas and geothermal fields. If you like we can discuss about this problem and we have a strong experience in data processing and seismic network. Enclosed you find a recent presentation about our experience with a particular focus on induced seismicity in geothermal fields and several years ago I processed seismic data recorded with the seismic network around oil/gas field near Malossa.

My office is in via Rubattino very closed to ENI head quarters.

Ciao

Fabio

Dear Lorenzo Pastone Barzaghi

I am really a gree with you that it seems to a waste of money and time to buy the instrument and to suitably prepare the site, because your target depths are 1-3 to monitore 5 * 10 km area, in actually any type of seismometers either short or broad band will be useful, but by using other topic of seismological study called Microtremor study, which is very inexpensive, frindly enviromentaly and can be used in any places, this new study will give you approximately clear and full seismological and geophysical and geotechnical imaging and monitoring to area from the surface through your target 1-3 km up to Conrad discontinuity and more that.

please if you have any interest to this new method, please do not hesitate to contact me to: [email protected]

Fabio,

I know about the guidelines and I am eager to see what will be the result. We do internally all the workflow for microseismic monitoring, from network design to result interpretation. However, it could be useful to arrange a meeting for you to present the skills of your team.

Ciao,

Lorenzo

Omar,

sorry, but we need to install seismic network for earthquake detection, so we have to go for this for sure.

About the technique you call Microtremor study, is it something related to low frequency passive seismic or passive seismic spectroscopy? If yes, we already tested it in the past as a direct hydrocarbon indicator with not conclusive results.

Ciao,

Lorenzo

Dear Lorenzo Pastore Barzaghi

The technique called Microtremor study is only the field work, which is mean we need only to make continous microtremor record by using any type of Microtremor device, which is a very low cost, and dosn't important if it has short or broad band seismometer, while the needed  time of field record not more than 17 min, this time enough after interprretation to extract directly all physical parameters due to all earth layers and it's contents, either to liquid or other earh materials ( solid materials ), and please note that the method ( New and Innovative Mathematical Theory in Seismology, Geophysical and Geotehcnical  ) is direct seismic and not passive, and note please that this method give direct results to explore underground water, Oil& Gas and of course all technical studies like the cavities, surface water, loose and solid material ( Please note by the thanks of Allah the all applicatios of this new theory gave excelent results, espicially for prospecting and exploring underground water, Oil& Gas,  Cavities and subsurface stifness materials ). Also this method give very big chance to study earh in globe, because we can deal to depth approximately up to 3200 km, where is Conrad discontinuity at depth about 14- 17 km very clearly in some pleases, while it is disappear in other places.

Dear Omar,

thank you for your explanation. Anyway, we need to install permament seismi network to monitor the presence or absence of earthquakes.

Best Regards,

Lorenzo

 Hi again Lorenzo!

Sorry for the delay. This is a hectic moment.

The second thing I was commenting is about lo-cut filtering and data conditioning.

If you want to design or constrain data this will be finite bandwidth. And having extended your data bandwidth will enhance stability (smoother slopes in your filters) and less risk. As per the aliasing the lo band wilk just a wittnes in fk to testify what signal content you have and thus discriminate any aliased high frequency (shallow environmental noise??), which will leak to the lowest frequencies.

While a broadband sensor can be useful for local earthquake studies, I'd be tempted to go for something shorter period than the 120s period you mention in your original post.  Something like a Guralp 6TD might be more than sufficient - much cheaper too!

Hello Bernal, now it's more clear. Thank you.

Dear Ian, I agree with you.

Dear Lorenzo,

you can save some money by installing short period seismic sensors at surface with the risk to have a low signal to noise ratio. With this choice, you cannot use standard detection techniques based on STA/LTA (short term versus long term average ratio). The human work cost will increase for processing noisy data, by using e.g. cross correlations and phase matched filters and searching sample by sample in the continuous waveforms repeaters of the largest well recorded microtremors. A good compromise could be to install downhole short period sensors (10-15 m ) to improve data quality. Take care that broad band sensors at surface (shallow soils or in presence of rock alteration) often are useless for detecting micro-nano seismicity with standard techniques.  

Very good point Alessandro.

A couple of papers on the subject that may be of interest. There are, of course, others.

https://www.researchgate.net/publication/259544213_Locating_microseismic_events_using_borehole_data

https://www.researchgate.net/publication/262573919_Characterization_of_fractures_and_faults_a_multi-component_passive_microseismic_study_from_the_Ekofisk_reservoir

Ian

Article Locating microseismic events using borehole data

Article Characterization of fractures and faults: A multi-component ...

Dear Ian,

sorry, the matter of seismic monitoring in deep boreholes is not the point here.

Dear Alessandro,

from your answer I understand that both short period and broad band sensors cannot be used at the surface aand need to be installed into shallow boreholes. But, as Fabio Moia pointed out earlier, the site preparation for borad band require special care. That is due to the higher noise recorded by broad band sensors. I agree that it is better to have short period sensors in shallow borehole and the whole cost  will be less than that for a braod band installation

You need to base your decision on the high frequency data you want to record.  For the 10 km range, and microseismicity you should be able to information above 200 Hz.  To be able to record that high frequency, you are likely looking at a 15 Hz exploration geophone, or perhaps a 4.5 Hz geophone.  The conventional 1.0 Hz geophones get unreliable above around 80 Hz.  Also, have you thought about what information of real use would come from frequencies below 1.0 Hz. 

Dear Lorenzo

you do not need a broad band seismometer for that area, it is a waste of money. Off course it depends what you want to study.  I suggest that you can use seismometer as the Lennartz 3D 1 Hz, but be sure that you usa a 24 bits Data Adquisition System

There are three types of information one can get from an array.  First location and source properties.  For these you need to record in a frequency band centered about the expected corner frequency.  For moment you need frequencies typically a factor of 10 below the corner frequency.  For slip properties you need frequencies a factor of 10 above the corner frequency.  The corner frequency is a function of fault length (and magnitude). For magnitude 2 the corner frequency is around 15 Hz.  However, in small arrays monitoring shallow seismic events you can expect to record Magnitude -3 with corner frequencies in the range of 100 to 200 Hz.   Second, inversion of reflections for crustal structure.  For most inversion applications, you are looking at the typical Oil exploration type geophones.  Third, if using the array to study structures much larger than the array or detect teleseismic events, then the long period geophones are appropriate.  In any case, the suggestion above to use  24 bits DAS is well taken because at these short distances your sensors will experience a wide variation in amplitude.   Also,the information gained from 3-component data will be superior to that obtained from single (vertical) component systems.

I thank all of you for the interesting answers to my question. I sum up what I understood till now:

• Of course, I assume 24 bits and 3C as standard.
• Very Broad Band (120s and similar) are useless, expensive and complex for my scope.
• Broad Band (30s-10s) can be a better option, mainly for seismic moment and processing stability.
• Anyway, I need high frequency, even greater than that of Short Period, for low magnitude, close earthquakes. With appropriate sampling, of course.
• Installation in shallow borehole to reduce ambient noise is a good option for Short Period and high frequency sensors.

Please, let me know other comments and suggestions.

Regards,

Lorenzo

Dear Lorenzo,

plaese give a look to our website http://rete-collalto.crs.inogs.it/en

OGS in Italy is monitoring (10 stations) the gas storage reservoir of Susegana in the Venetian foothills (Edison). If you are interested in technical details of our seismic network send an email to [email protected] (stations are in boreholes, one of them  down to 100 m and are operating since 2012.

Dear Alessandro,

now I understand what is going on.

I saw that in the Collalto network you installed the seismic station ED06 with a 120s seismometer, at 5 m depth. You kept it working for a couple of years and then changed the sensor to a 30s seismometer. So, you probably realized that such a VBB sensor is useless to monitor microseismic activity for a reservoir at about 1500 m of depth. Hence, I am comforted from your experience which is the best answer to my original question. Probably, not all are aware of this eveolution of your approach.

Last question: is the operator of the field aware that he paid for a VBB sensor that is non more working in its network?

Ciao,

Lorenzo

Dear Lorenzo 

I am not involved in the Collalto reservoir monitoring please contact the responsible Dr. Enrico Priolo ([email protected]) or for technical details Dr. Marco Romanelli ([email protected]). They can easily share with you technical details and problems encountered especially in maintaining downhole sensors. 

ciao

Alessandro

May i use this medium to pose this question: which sensor is most appropriate for local seismicity monitoring, site effects investigation and local tomography? We are having difficulties recording local events with our broadband sensors.

Regards.

Umar

If you are looking at the cause, stress drop, and mechanism of small reservoir induced events you need the high-frequency response. Typically a factor of 10 in frequency above the corner frequency.  That can approach 200 Hz for mag. 0 to -3 events. The longer periods those longer than 1.0 s may be needed only for site response or for looking at teleseismic data.

I recomend to use the Lennart MkIII 1Hz (3D) Is ideal local and regional studies.

we ha e it in our regional network (RESAJ)

Dear Umar,

for monitoring local seismicity useful frequency band 1-100 Hz is ok. If you want to be conservative you can enlarge a little bit the low frequency band down to 5-10s, no more.

I agree with Francisco Lennartz Le-3D1s is ok, you can use also Lennartz Le-3D5s, or mixing the two type of sensors.