Detector Development and Cost Optimization at SiD: Insights and Progress

SiD DBD Optimization & Costs M. Breidenbach 15 November 2011 SiD DBD SLAC 1

Outline The Optimization Process Schedule?? CLIC Coordination on costs Costing cultures Effect of detector variations. Cost Sensitivity to selected commodity inputs. First Steps 15 November 2011 SiD DBD SLAC 2

Detector Development Process Physics Motivation Benchmark Reactions Required Detector and Accelerator Performance Detector Design Concept Parametric Variation with Costs Evaluate Technical Realizability of Concept Simulation of Physics and Backgrounds Research & Development of Necessary Technology Reconstruction Engineering Evaluation & Modification of Concept Analysis: Signal & Background Evaluation Rational Design? Benchmark Performance vs Cost 15 November 2011 SiD DBD SLAC 3

Major Ongoing Program: Optimize design maximize performance vs cost Select global parameters: Radius, z extent & number of layers in Tracker Seems to be stable. Depth of hadron calorimeter in I No Progress Strength and uniformity of solenoidal field Seems to be stable Evaluate performance for more subtle issues such as un-instrumented support or readout structures & calorimeter cracks. Requires: Physics event generators Background generators Simulation & Reconstruction framework PFA Algorithms Analysis Detector Costing Tools 15 November 2011 SiD DBD SLAC 4

Detector Technology R&D The ILC Interaction Point environment is benign compared to the LHC. The detectors have only trivial rate and radiation requirements. However, the required measurement capability is, in varying degrees, beyond the state of the detector builder s art. SiD is developing detector architecture and working on strategic R&D to show that these proposed approaches are viable. 15 November 2011 SiD DBD SLAC 5

Developing A Particle Flow EM Calorimeter Segmentation, Segmentation, Segmentation Kapton Cables Pixel Sensor with KPiX 1 mm readout gap 13 mm effective Moliere radius. Very compact showers Problems with bump bonding. Slow Progress. 15 November 2011 SiD DBD SLAC 6

Parametric Costing - LOI SiD has developed a tool which can consistently vary global detector parameters and generate M&S and Labor costs. Vary Rtrkr, B E/E = 3.8% SiD Base M&S Cost vs Tracker Radius 15 November 2011 SiD DBD SLAC 7

Physics Performance vs Cost - LOI Fractional error in measurement of the triple Higgs coupling vs cost. Each point is a separately optimized detector with different PFA E/E 15 November 2011 (Cost is ILC units (M&S only) SiD DBD SLAC 8

CLIC ILD Agreed Unit Costs 15 November 2011 SiD DBD SLAC 9

Superconducting Solenoid Superconducting coils of this scale are difficult to estimate. Usual practice is to scale by stored energy, either a + bE or cE0.66 Attempt was made to extract the CMS cost for cold mass and cryostat - ~$48M. Obtained industrial estimate for SiD coil; ~same cost; but the stored energy! SC Coil Costs 140.0 120.0 100.0 Linear 80.0 M$ (2005) Power 60.0 40.0 20.0 0.0 0 0.5 1 1.5 2 2.5 3 3.5 4 Stored Energy (GJ) 15 November 2011 SiD DBD SLAC 10

Superconducting Coil, continued We have chosen to be conservative, and are using the linear fit to Babar at the low end and the industrial estimate to get the slope. SiD is pursuing R&D on an advanced conductor that would be significantly simpler than the CMS conductor. Extensive discussions with ILC Detector Cost Group and CLIC. Conclusion: ILD will stay with in house construction (CERN, Saclay, other) SiD will stay with industrial production to a performance specification. ILD has lower cost, owns the risk. SiD has higher cost, pays industry to take some risk. Difference hard to quantify because ILD does not include vacuum vessel with solenoid, perhaps $20M. 15 November 2011 SiD DBD SLAC 11

Platforms After long haggling, SiD has agreed to platforms for push- pull providing: SiD specifies functional parameters, particularly area, elevation to Beamline, and vibrational performance. ILC pays for it! Makes some sense for vertical access site. Still troubled by concept for horizontal access (Japanese) site. 15 November 2011 SiD DBD SLAC 12

Parametric Model Self consistent Excel model of SiD e.g. tracking layers and disks adjusted to fit allocated space Calorimeters adjusted to nest properly with the tracker Solenoid is adjusted for its radius and field Iron is adjusted to return the flux. Fundamental parameters can be varied: e.g Tracker radius and aspect ratio; N layers Calorimeter N layers, thickness, materials, gaps Costs calculated for each system: Cost driving component counts are calculated e.g. tungsten plate area, silicon detectors, and KPiX for the EMCal Model has cost tables for these M&S items and associated tech labor in hours Costs that are ~fixed, e.g. engineering or fixturing, are imported from the Work Breakdown Structure. Macros allow easy variation of parameters to calculate cost derivatives. 15 November 2011 SiD DBD SLAC 13

Work Breakdown Structure Hierarchical structure breaking down SiD to recognizable and understandable units. Separate tables for purchased M&S and labor. Contingencies for each item, propagated through the WBS. Cost are estimated in 2008 US$. Labor both costed in $ and summed by man-years. Labor is estimated in ~50 different types, e.g. Project Engineer, plumber, iron worker and condensed to the ILC categories Engineering, Technical, and Administrative. The ILC style cost is defined here as the base M&S cost without contingencies, plus the labor in man-years in the 3 categories. 15 November 2011 SiD DBD SLAC 14

SiD w New Unit Costs (4.5 Hcal) ILC Costs M&S Administrative (MY) M&S Base (M$) Contingency (M$) Engineering (MY) Technical (MY) 1.1.1 Beamline Systems $ 3.68 $ 1.42 4.0 10.0 0.0 1.1.2 VXD $ 2.80 $ 2.04 8.0 17.7 0.0 1.1.3 Tracker $ 17.74 $ 6.87 24.0 53.2 0.0 1.1.4 EMCal $ 99.86 $ 39.87 13.0 287.8 0.0 1.1.5 Hcal $ 50.18 $ 19.54 13.0 27.2 0.0 1.1.6 Muon Sys $ 5.28 $ 1.63 5.0 19.8 0.0 1.1.7 Electronics $ 4.90 $ 1.65 44.1 41.7 0.0 1.1.8 Magnet $ 112.53 $ 38.66 29.2 25.0 0.0 1.1.9 Installation $ 4.10 $ 1.08 4.5 46.0 0.0 1.1.10 Management $ 0.92 $ 0.17 42.0 18.0 30.0 Totals $ 302 $ 113 187 546 30 15 November 2011 SiD DBD SLAC 15

Cost Systems US: Cost everything! Convert labor years to $$ Europe: Contingency uncertain, time is contingency . Usually labor not included. Escalation usually not estimated. 15 November 2011 SiD DBD SLAC 16

Cost Systems M&S Labor Totals $302 $81 $383 Base $113 $28 $141 Contingency $415 $110 $525 Total 0.06 0.20 Indirect rates $25 $22 $47 Indirects $440 $132 $571 Totals w indirects 2008 Total in FYXXXX M$ 571.3 2016 Start Year 6 years Construction Duration 1.035 per year. Inflation 1.460 Factor 262.8 Total Escalation 834.1 Total, TYM$ ~x 2.8 15 November 2011 SiD DBD SLAC 17

Costs by Type SiD Costs by type M&S $302 Labor $81 Contingency $141 Indirects $47 Escalation 262.8 Total $834 15 November 2011 SiD DBD SLAC 18

Subsystem M&S SiD M&S Costs by Subsystem $120,000,000 $100,000,000 $80,000,000 Cost $60,000,000 $40,000,000 $20,000,000 $- Subsystem 15 November 2011 SiD DBD SLAC 19

SiD Labor by Subsystem SiD Labor 350.0 300.0 250.0 Man-Years Engineering 200.0 Technical 150.0 Clerical 100.0 50.0 0.0 EMC al Magnet Hcal VXD Installation Management Tracker Muon Sys Electronics Beamline Systems Subsystem 15 November 2011 SiD DBD SLAC 20

M&S Cost vs Hcal Thickness HCal Thickness 390 370 350 330 M$ 310 290 270 250 3.0 4.0 5.0 6.0 7.0 HCal Lamda 15 November 2011 SiD DBD SLAC 21

M&S Cost vs Hcal Gap HCal Gap 335 330 325 320 315 310 305 300 295 290 285 0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 15 November 2011 SiD DBD SLAC 22

Cost Sensitivity Effect of doubling the nominal unit cost Item Nominal Unit Cost SiD Base M&S Cost (M$) Magnet Iron (finished and delivered) $6/Kg 48 Tungsten (powder alloy) plate $180/Kg 14 Si Detector $6/cm2 77 HCal Detector $2000/m2 7 15 November 2011 SiD DBD SLAC 23

Cost Comments I This estimate is adequate for a Letter of Intent. Not much more is likely to be done for the DBD. It is not adequate for a DOE Project Review (Lehman) Adding contingency would be advised when thinking about these numbers. Many commodity costs remain unstable but we now use many of the same numbers as CLIC. 15 November 2011 SiD DBD SLAC 24

Cost Comments II SiD now stands at $302M Base + $113M Contingency + 760 MY ILD (LOI) is 407 MILCU + 1400 MY. The nominal model for funding is in-kind contributions from the governments about the same model as for ILC. ILC is strenuously working to reduce and control costs. There have been some questions regarding the cost difference, with the inference that ILD is a better performing detector. Is this true? For what physics? 15 November 2011 SiD DBD SLAC 25

SiD Next Steps Set HCal thickness. This decisions should be based on physics performance! Need to evaluate new 1 TeV benchmark performance, and think about performance comparisons to ILD. With the HCal parameters set, we can do a small iteration on the detector mechanics, and have a better model of cracks and dead areas. The DBD will not be a TDR!!! 15 November 2011 SiD DBD SLAC 26

The Bottom Line Actually building SiD would take 5-6 years from a TDR and funding. It will take 3 1 years and funding to produce a TDR, which will include: All technology choices except (perhaps) the VXD. Beam and B field tested prototype detectors. Production Prototype detectors Complete electronics designs with fully functional prototypes Reviewable plans for actual design, management, costs, and funding. ILC appears to be headed for a TDR-lite. The DBD is not a TDR! 15 November 2011 SiD DBD SLAC 27

Backup Slides 15 November 2011 SiD DBD SLAC 28

WBS 15 November 2011 SiD DBD SLAC 29

US Costs (if you are curious) M&S Labor Totals $271 $82 $353 Base $101 $28 $129 Contingency $372 $110 $482 Total 0.06 0.20 Indirect rates $22 $22 $44 Indirects $394 $132 $526 Totals w indirects 2008 Total in FYXXXX M$ 525.9 2016 Start Year 6 years Construction Duration 1.035 per year. Inflation 1.460 Factor 241.9 Total Escalation 767.9 Total, TYM$ 15 November 2011 SiD DBD SLAC 30

U.S. DOE style costing Contingencies are assigned to M&S and labor Allows extra funding to hold a schedule in the face of unforeseen problems. Fund items that were missed in the estimate Provide some relief from under-estimates Transform labor to $ value using SLAC salary numbers including benefits, but not overhead. Compute indirects as fraction of M&S and Labor. SLAC large Project numbers are used. Escalation (inflation) calculated assuming a start date, a 6 year construction cycle, and an inflation rate. Assume 2016 start, inflation = 3.5%/year These assumptions are uncertain The escalation is substantial 15 November 2011 SiD DBD SLAC 31

Interface Assumptions IR Hall, with finished floor & walls, lighting, and HVAC are provided by ILC. Utilities, including 480 VAC power, LCW, compressed air, and internet, are provided on the hall wall. External He compressor system with piping to the hall is provided. The refrigeration and associated piping is an SiD cost. Any surface buildings, gantry cranes, and hall cranes are provided by ILC. Data storage and offline computing are provided by others. Detector motion rails (both for push-pull and detector opening in beamline and garage positions) are installed by SiD in suitable channels provided by ILC. 15 November 2011 SiD DBD SLAC 32