Spsc protodune-sp preliminary tdr review
Yüklə 202,77 Kb.
|
- Bu səhifədəki naviqasiya:
- Major Comments: 1) In the view of the referees
- Overview of the ProtoDUNE-SP Programme
- Optimal ProtoDUNE-SP programme 2018
- 3) In addition, it is mandatory to estimate what is the schedule and manpower contingency that is affordable for
- 5) The cold electronics are the critical system at this time. The TDR should be updated with the results of the cold electronics review and testing underway at FNAL and BNL before the annual review.
- This update should also address the following points. (i) What fraction of the electronics must be active to take meaningful data
- (ii) There are possible single-point failures in the electronics components. A recovery plan should be included that describes what will happen in this event, e.g. if a group of wires is lost
- Technical Comments: 1 Introduction
- 1.3. The statement is made that the second goal is for
|
SPSC ProtoDUNE-SP Preliminary TDR Review 27 January 2017 The SPSC referees have reviewed the preliminary Single Phase ProtoDUNE (ProtoDUNE-SP hereafter) TDR version submitted to the SPSC at the October 2016 meeting. This document contains the referee comments, organized as major, technical and clarification comments. We request the collaboration to address the major and technical comments in a stand-alone written document, by February 27. We will schedule a phone meeting with representatives of ProtoDUNE-SP and the referees to follow up outstanding questions during the week of March 6. The referees will deliver their final report to the SPSC Chair by March 15.
Major Comments: 1) In the view of the referees, the major risk of the project is the schedule to take beam data before LS2. As input to the CERN MTP, the proponents should address how they plan to mitigate the risk of (i) a shortened data run with beam before LS2, and (ii) no beam before LS2. Overview of the ProtoDUNE-SP Programme ProtoDUNE-SP has four main aims, all of which are essential parts of the DUNE far detector development programme:
At this point in time, the protoDUNE-SP construction project remains on schedule for being ready for beam operation in July 2018. However, there is little schedule contingency in the construction project. There are three main risks to the ability of protoDUNE-SP to achieve its physics goals in during the beam campaign in 2018:
For these reasons, at this stage, we do not wish to exclude the possibility of beam operation after LS2. Independent of the scientific case for the test beam campaign, we believe there is a very strong argument for operating the protoDUNE-SP (and protoDUNE-DP) detectors for as long as possible prior to DUNE construction. The liquid argon technology is sufficiently new that long-term testing of the prototypes is highly desirable to reduce the risks of problems arising during the operation of the DUNE far detector modules. Once the far detector modules are installed, they need to operate reliably for 15-20 years. The protoDUNE detectors can be used to mitigate some of the risks associated with long-term operation. This argues for cosmic-ray operation in 2019 and, assuming the availability of resources, also in 2020. We believe this is an important part of our risk mitigation strategy for the major DUNE detector construction. Optimal ProtoDUNE-SP programme
Answers to the four related points raised by the SPSC referees are given below. i) what is the duration and resources needed for cosmics running during LS2? We believe there is a strong argument for operating the protoDUNE-SP (and protoDUNE-DP) detectors for as long as possible prior to DUNE construction. The liquid argon technology is still sufficiently new that a long-term test of the prototypes is an important part of our risk mitigation strategy for the DUNE construction. For this reason, we believe there is a strong case for cosmic-ray operation in 2019 and ideally 2020. The technical resources required for operation in this mode are being understood. For cosmic-ray operation, the data storage requirements will be relatively low as we would only record a fraction of the potential trigger rate. We have discussed the likely needs for CERN technical support with the leader of the Neutrino Platform. For cosmic-ray operation during LS2 it is estimated that 0.2 FTEs of CERN technical effort are required for protoDUNE-SP specific cryogenic support. The additional consumables cost mostly will be related to the consumption of the nitrogen necessary to keep the LAr cold, estimated to be approximately 300 kCHF/year.
The critical project decisions for the first two far detector modules will be taken in 2019, at which time DUNE plans to produce TDRs for international peer review (LBNC) and for the DOE CD-2/3B review scheduled for October 2019. By this time, we need to have validated that the detector elements installed in protoDUNE-SP meet the Far Detector requirements; this can be achieved using cosmic-ray muon tracks. The beam programme is needed for the physics calibration of the response of the detector to different particle species and this is independent of the basic validation of the performance. Beyond the TDR, there will be a period of about one year before the Production Readiness Reviews that launch the construction of the elements of the far detector. Operation of protoDUNE-SP during 2019 (and ideally 2020) would provide further confidence in the long-term stability of the design. By 2021 we will be constructing the elements of the first detector module and would be starting the procurement of the materials for the second detector module. At this stage, we wish to keep open the option of recording test-beam data after LS2. This may be necessary if there are delays to the schedule in the installation of either the detector and/or cryogenic systems or if there are issues with the quality of the beam during operation in 2018. However, this operation to record physics calibration data would not have an impact on the timing of DUNE critical decisions. DUNE will build four 10-kt far detector modules. Our strategy is for the first module to be the single-phase design as prototyped using protoDUNE-SP. We are not assuming that the subsequent modules will be the same design. For example, ProtoDUNE-DP prototypes an option for second and/or subsequent far detector modules. There may also be evolutions in the single-phase design. Consequently, even assuming a successful test-beam campaign in 2018, we would like to keep open the possibility of additional large-scale prototyping beyond LS2. This would allow prototyping of improvements to the detector technology for the third and fourth DUNE far detector modules. For example, if there were a new generation of cold readout electronics the physics calibration could be established with the existing TPC. The incremental costs would be relatively small compared to the initial investment in the facility.
The [CERN] resources required for decommissioning would depend on the scope of the decommissioning activities. For example, simply removing the TPC from the cryostat is a much smaller effort than removing the cryostats and cryogenic infrastructure. We are not in a position to provide a reliable estimate at this stage. The Neutrino Platform team are best placed to estimate the decommissioning needs. iv) the impact on the overall DUNE schedule in case an additional 1y run is performed after LS2 to complete the ProtoDUNE physics programme. Assuming that the basic design of the detector has been validated using cosmic-ray muons, there would be no impact on the DUNE schedule. The beam data is required to understand in detail the detector response to different particle species. It will also be an essential component in tuning simulation models, for example for pion-argon interaction cross sections. The beam data is thus needed for the DUNE physics programme, but not for the validation of the detector. 2) The TDR gives a good general overview of ProtoDUNE-SP and indicates that the most critical issues were properly identified. In normal conditions, this would be a very good starting point to be integrated with sharing of responsibilities, procurement schedule, schedule of installation and Q/A tests. In the specific case of ProtoDUNE-SP, given the tight schedule constraint due to LS2, (i) procurement, (ii) installation schedule, and (iii) Q/A schedule play a crucial role and detailed documents should be provided before the Annual Review in April 2017. These documents will be provided prior to the Annual Review. 3) In addition, it is mandatory to estimate what is the schedule and manpower contingency that is affordable for:
We accept that there is little schedule contingency on the critical path. We have not performed a risk-based assessment of the likelihood of a delay in the schedule - this would be dominated by unexpected issues with sub-systems. However, a detailed risk assessment has been performed and risk mitigation strategies have been put in place to reduce the likelihood of the realization of risks that could impact the schedule. For example, the schedule risks associated with uncertainties in the duration/complexity of the installation procedure have been mitigated by the full-scale integration test at the Ash River site. Technical risks, which if realized, would impact the schedule are being mitigated by targeted testing programs. For example, the CPA/HV/FC system will be tested at Fermilab in the 35t cryostat. An extensive plan for testing/QA of the cold electronics has been developed. For the specific questions:
4) Given the experience of the 35t cryostat and MicroBooNE, it would be highly advisable to have a recovery plan in place if major noise issues arise after the installation of the cold electronics (CE), at least for an identified design fault in the cold TPC front-end electronics. MicroBooNE has been taking data for 1.5 years in stable conditions with acceptable levels of noise. The LBNE 35t prototype was initially intended as a demonstrator for reaching the necessary argon purity (electron lifetime), and the subsequent installation of a TPC in a cryostat, which was not designed for this purpose, contributed in part to the noise issues. The lessons learned from this experience have been taken into account in the ground and noise mitigation measures for protoDUNE-SP. Our strategy is to pursue a detailed testing program for the electronics and system level tests in the cold, prior to installation in the cryostat. 5) The cold electronics are the critical system at this time. The TDR should be updated with the results of the cold electronics review and testing underway at FNAL and BNL before the annual review. We plan to update the TDR to incorporate new information, but the full report goes beyond the scope of the TDR. We are happy to supply the SPSC with additional documentation. A new task force is currently performing an alternatives analysis on the cold electronics design for DUNE. This update should also address the following points. (i) What fraction of the electronics must be active to take meaningful data? It is difficult to quantify/define “meaningful data”. With three readout wire planes, there is some redundancy. Unusable parts of the detector will be defined by overlapping regions of two inactive wire planes. Even 10 % inactive channels (far larger than our goal) would correspond to a 1 % dead region of the detector (i.e. only one readout plane viewing this region). (ii) There are possible single-point failures in the electronics components. A recovery plan should be included that describes what will happen in this event, e.g. if a group of wires is lost? ProtoDUNE-SP is a prototype. One of the main aims is to establish that the TPC design meets the detector requirements. This does not require that all wires are active. Experience with the analysis of MicroBooNE data shows that locally inactive regions (groups of dead wires) can be handled using the redundancy in the three-plane read out. If 1% of channels were inactive, this would have minimal impact on the protoDUNE-SP program. Even if 10% of the channels were inactive, it would most likely be possible to establish that the TPC design meets the detector requirements. (iii) What are the circumstances under which electronics components would be replaced? (iv) what performance goals have been validated in prototypes? It is often unclear in the TDR whether design decisions are based on test results or simulations. The modular nature of the electronics means that it is possible to make a replacement, but this would be a significant operation that would require emptying the cryostat and reinstalling the electronics with difficult access limitations. We have not worked through a plan for the replacement of the electronics; we do not believe this is the most efficient use of valuable engineering resources - all efforts are being directed towards risk mitigation through testing and QA/QC. 6) Given the DAQ maximum operation rate, the muon halo should be estimated fully to understand its impact. Can meaningful physics data be acquired with the DAQ as designed and the muon halo estimated from a full calculation, including the target 37m from ProtoDUNE-SP and the neighboring beam lines, in addition to the primary production target? Is active shielding mitigation required? Beam triggers from the beam counters along the ProtoDUNE-SP dedicated tertiary beam line will feed the DAQ up to its 25 Hz max operating rate. The muon halo (from the decay in flight of the H4 secondary high energy pion beam) has no impact on the DAQ. Pile up in the recorded beam event frame (2ms drift time) from cosmics and muon halo should not prevent full identification of the incident beam particle and complete reconstruction of the event and its interaction process due to the 3D imaging capability of a LArTPC. This capability has been demonstrated by MicroBooNE, which has been able to identify neutrino interactions in the presence of cosmic ray pile up. ProtoDUNE-SP presents an easy problem as the beam particle entry point and time is known. The flux of punch-through particles originating from incident pion beam interactions in the secondary target at 37 m from the detector will be strongly reduced by passive shielding around the target and at the end of the H4 tertiary beamline. Radiation shielding design assessment is currently under development through detailed MC simulations (CERN RP and Neutrino Platform). 7) The total data volume reaches 3 PByte in a few weeks, a substantial fraction of yearly LHC rates. What are the resources planned for use to handle these large data volumes, in particular the CERN resource? This should be documented in the TDR. The estimated 3 PByte is the anticipated total raw data volume to be taken during the planned SPS run. It is based on a trigger rate of 25 Hz and a number of conservative assumptions, for example the data reduction from lossless compression. If absolutely necessary, there are ways to reduce the data volume, either through lossy compression or by simply downscaling the trigger rate. The computing provision for protoDUNE-SP is being discussed by the dedicated CERN/Fermilab/DUNE interface group, which consists of stakeholders from protoDUNE-SP, protoDUNE-DP, DUNE computing, CERN IT and the FNAL computing sector. The current understanding is that storage of the raw data itself is not a major issue. Adequate resources for the transfer the data from CERN to FNAL, storage of data at FNAL, and processing the data at FNAL and elsewhere are being planned by this group. Technical Comments: 1 Introduction 1.3. What is the proposed run duration to meet the goals of “Operating the detector in real experimental conditions and for an extended period will allow for a full characterization of the components, including the membrane cryostat and the cooling and purification circuit, the APA design and the layout of its cold read-out electronics, the HV system, and the PDS and its warm read-out electronics.” Commissioning and successfully bringing the detector into operation and maintaining it in stable operating conditions over the order of (1-3) months would allow for validation of the design from the perspective of basic detector performance – this can be achieved by collecting cosmic-ray data. The large samples of test-beam data of different particle species would allow for a more in-depth understanding the response of the detector and ultimately be used for calibration purposes. A run period of approximately three months in the test beam is considered to be sufficient, assuming stable operation under good beam conditions. The aim is to accumulate these data before the start of LS2 (Jul-Oct, 2018). The liquid argon technology is, however, sufficiently new that a long-term test of the ProtoDUNE prototype, extending beyond the start of LS2, would be highly desirable to reduce risks associated extended operation of the DUNE far detector modules. This argues for cosmic-ray operation in 2019 and most likely 2020. We believe this extended ProtoDUNE running would be an important component of our risk mitigation strategy for the overall DUNE detector construction effort. 1.3. The statement is made that the second goal is for “ProtoDUNE-SP [to] perform the measurements needed to understand, and quantify to the extent possible, the systematic uncertainties that will affect the DUNE oscillation measurements.” The TDR should specify which oscillation measurement systematics the beam test measurements constrain, at what level, and connect the required quantitative constraint with the proposed run duration. The run duration is primarily constrained by the start of LS2. Three types of systematic effects will be addressed: i) a detailed understanding of energy scale and energy resolution for electromagnetic and hadronic showers; ii) the detector response to different particle species, including measuring the impact of recombination effects as a function of angle; and iii) improved modelling of the interactions of hadrons in Argon, which will provide constraints to the GEANT4 physics models. We believe a data set of order 10M beam triggers (hadron and electron beam combined) will be sufficient for this purpose - it will certainly represent a step change in our understanding of the effects listed above. The data samples listed in Tables 1.1 and 1.2 of the TDR will provide of order 100k events per energy point for each event sample. This will allow potential systematic biases to be probed at the 0.5 % level in each sample. Making a more quantitative statement regarding the level at which systematics in DUNE oscillation measurements can be constrained through analysis of ProtoDUNE test beam is not possible at this time. There are ongoing detailed studies dedicated to this topic. The aim is to have a solid answer by the start of the test beam run in 2018, which will allow us to modify the exact strategy for the run plan (if needed). What is the risk to the DUNE oscillation analysis of acquiring only cosmics data in the case of schedule slippage? Related to this point, a number of statements in this section (in particular, the last few paragraphs) depend on how well the reconstruction works, e.g. success of calorimetric energy reconstruction, and are strongly dependent on the hadronic interaction model. To what extent this data is useful depends on systematics, modelling, reconstruction, etc. What are the assumptions that have been used here? The beam time request should include a description of the analysis that has been done, and what the systematics assumptions are, for determining what number of events is required to meet the DUNE oscillation analysis physics goals. There is little schedule contingency on the critical path of the construction project. There are also risks associated with the schedule for the installation/commissioning of the cryogenic systems and the availability/quality of the beam line. Consequently, there is a risk that the desired test-beam data samples, needed for protoDUNE-SP to achieve its physics goals, will not be accumulated before LS2. Therefore, we do not wish to exclude the possibility of beam operation after LS2; we believe this data is necessary to address exactly the issues raised here. The analysis of the test-beam data will involve strong feedback between reconstruction, detector simulation, and hadronic modelling with the ultimate goal of achieving agreement at the sub-percent level. In principle, this can be achieved with data samples of order 100k cleanly identified particles at each energy point. If we can reach this level of agreement, we will be able to constrain detector systematics on the energy scale of neutrino interactions to the level of 1%. We believe the event samples in the TDR run plan are sufficient for this purpose. In the absence of LS2, we would request a longer run, but not one of an order of magnitude larger. Determining the precise number of beam events that are required to meet the DUNE oscillation analysis physics goals is currently under study - but this it is not driving our beam time request. It should be noted that the run duration will heavily depend upon actual detector performance and the beam quality (precision in particle momentum, impact position, beam PID) and until the commissioning of both LArTPC detector and H4 beam-line, there will be large uncertainties. For these reasons, we do not wish to exclude beam operation after LS2. Yüklə 202,77 Kb. Dostları ilə paylaş:
|