1 00:00:01,640 --> 00:00:04,040 The following content is provided under a Creative 2 00:00:04,040 --> 00:00:05,580 Commons license. 3 00:00:05,580 --> 00:00:07,880 Your support will help MIT OpenCourseWare 4 00:00:07,880 --> 00:00:12,270 continue to offer high quality educational resources for free. 5 00:00:12,270 --> 00:00:14,870 To make a donation or view additional materials 6 00:00:14,870 --> 00:00:18,830 from hundreds of MIT courses, visit MIT OpenCourseWare 7 00:00:18,830 --> 00:00:20,000 at ocw.mit.edu. 8 00:00:22,940 --> 00:00:26,090 NATHAN PHILLIPS: Introductions, Audrey Schulman. 9 00:00:26,090 --> 00:00:26,990 AUDREY SCHULMAN: Hey. 10 00:00:26,990 --> 00:00:28,910 NATHAN PHILLIPS: HEET, Home Energy Efficiency Team. 11 00:00:28,910 --> 00:00:29,870 I'm Nathan Phillips. 12 00:00:29,870 --> 00:00:32,330 I teach at Boston University. 13 00:00:32,330 --> 00:00:36,440 And the previous speakers, I think, 14 00:00:36,440 --> 00:00:40,010 really talk to how the environmental problems 15 00:00:40,010 --> 00:00:44,360 that we have are either invisible or very hard 16 00:00:44,360 --> 00:00:45,120 to grasp. 17 00:00:45,120 --> 00:00:47,330 So being able to see these things with evidence, 18 00:00:47,330 --> 00:00:51,770 photographs or having simple and precise metrics 19 00:00:51,770 --> 00:00:56,360 to understand the magnitude of our climate impacts 20 00:00:56,360 --> 00:00:57,030 is important. 21 00:00:57,030 --> 00:00:59,900 So this is very much in line with this. 22 00:00:59,900 --> 00:01:03,020 And so some of you here have been with us 23 00:01:03,020 --> 00:01:07,160 from the start, when we met here a couple of weeks ago. 24 00:01:07,160 --> 00:01:10,440 Some of you came in midstream. 25 00:01:10,440 --> 00:01:15,020 So we laid out this problem about the methane gas leaks 26 00:01:15,020 --> 00:01:17,090 from the natural gas pipeline infrastructure 27 00:01:17,090 --> 00:01:20,000 under Boston, Cambridge, Somerville, Eastern United 28 00:01:20,000 --> 00:01:21,650 States. 29 00:01:21,650 --> 00:01:24,590 Then we talked about solutions. 30 00:01:24,590 --> 00:01:28,880 We had a hackathon about a week and a half ago or a week ago. 31 00:01:28,880 --> 00:01:32,390 And then, yesterday, Audrey and I, 32 00:01:32,390 --> 00:01:35,480 and some of the people in this room, hopped in a van, 33 00:01:35,480 --> 00:01:38,420 and we started sniffing these gas leaks 34 00:01:38,420 --> 00:01:40,970 and mapping them on Google Earth. 35 00:01:40,970 --> 00:01:44,180 And so we just wanted to share with you-- 36 00:01:44,180 --> 00:01:46,490 how many people were actually in the van with us? 37 00:01:46,490 --> 00:01:49,730 So a good chunk of us here. 38 00:01:49,730 --> 00:01:51,830 So this was us yesterday. 39 00:01:51,830 --> 00:01:57,019 And this is a combined map here of both of the trips 40 00:01:57,019 --> 00:01:57,560 that we took. 41 00:01:57,560 --> 00:02:02,060 So the first van load, we had a full van of 12 people, 42 00:02:02,060 --> 00:02:04,850 and we drove from here to Somerville and back. 43 00:02:04,850 --> 00:02:07,110 And then we pretty much did the same thing 44 00:02:07,110 --> 00:02:09,710 with a little slight variation in the second trip. 45 00:02:09,710 --> 00:02:14,450 So let's see-- so that's where we are right now, 46 00:02:14,450 --> 00:02:15,606 is that correct, Audrey? 47 00:02:15,606 --> 00:02:17,480 AUDREY SCHULMAN: Yeah, and an important thing 48 00:02:17,480 --> 00:02:19,390 to know is that there's a baseline about two 49 00:02:19,390 --> 00:02:23,285 parts per million in this, so that the wall, everywhere, 50 00:02:23,285 --> 00:02:25,035 is just a baseline of methane that exists. 51 00:02:28,364 --> 00:02:30,280 That used to be lower-- 52 00:02:30,280 --> 00:02:32,200 NATHAN PHILLIPS: No, let's share this, please. 53 00:02:32,200 --> 00:02:34,020 But actually please-- 54 00:02:34,020 --> 00:02:36,620 AUDREY SCHULMAN: Sorry? 55 00:02:36,620 --> 00:02:39,733 AUDIENCE: So it's Mercer Street, that's the methane leak? 56 00:02:39,733 --> 00:02:43,290 AUDREY SCHULMAN: Yeah, so there's a big league 57 00:02:43,290 --> 00:02:47,100 somewhere here, right in front of the most iconic part of MIT. 58 00:02:47,100 --> 00:02:48,420 AUDIENCE: [INAUDIBLE] 59 00:02:48,420 --> 00:02:50,170 AUDREY SCHULMAN: Oh sorry, you want me to? 60 00:02:50,170 --> 00:02:51,503 NATHAN PHILLIPS: Oh, I'll do it. 61 00:02:53,190 --> 00:02:56,920 AUDREY SCHULMAN: But the wall here that you see of everywhere 62 00:02:56,920 --> 00:02:58,110 where we drove? 63 00:02:58,110 --> 00:03:01,140 That's sort of just the parts per million of methane 64 00:03:01,140 --> 00:03:02,670 that's just background noise. 65 00:03:05,190 --> 00:03:08,085 And the big spikes are where there were leaks. 66 00:03:10,950 --> 00:03:17,130 So and what's interesting is, we did a US survey of Cambridge 67 00:03:17,130 --> 00:03:21,840 and Somerville two years ago, and there was a leak here. 68 00:03:21,840 --> 00:03:23,910 So there still is. 69 00:03:23,910 --> 00:03:26,150 It's nice to know. 70 00:03:26,150 --> 00:03:29,650 What do you want to say? 71 00:03:29,650 --> 00:03:33,780 NATHAN PHILLIPS: Well, how about we go to the area, 72 00:03:33,780 --> 00:03:35,840 where it was it, Prince Street in Somerville? 73 00:03:35,840 --> 00:03:36,870 AUDREY SCHULMAN: Pearl Street. 74 00:03:36,870 --> 00:03:38,120 NATHAN PHILLIPS: Pearl Street. 75 00:03:38,120 --> 00:03:39,320 Let's take a look at. 76 00:03:39,320 --> 00:03:43,170 AUDREY SCHULMAN: So here's us going by Kendall. 77 00:03:50,580 --> 00:03:52,050 And then back-- 78 00:03:54,942 --> 00:03:55,900 I'm having a prankster. 79 00:03:55,900 --> 00:03:57,733 NATHAN PHILLIPS: Thanks for driving, Audrey. 80 00:03:57,733 --> 00:04:00,172 I'm not good at Mac. 81 00:04:00,172 --> 00:04:01,380 AUDREY SCHULMAN: OK, yeah no. 82 00:04:01,380 --> 00:04:04,290 Mac is my language. 83 00:04:04,290 --> 00:04:10,430 So that's Pearl Street there, the big mountains. 84 00:04:10,430 --> 00:04:11,420 NATHAN PHILLIPS: Now. 85 00:04:11,420 --> 00:04:13,770 These are all National Grid, right? 86 00:04:13,770 --> 00:04:17,160 AUDREY SCHULMAN: Yeah so the difference between Eversource 87 00:04:17,160 --> 00:04:21,279 and National Grid is this train track right here. 88 00:04:21,279 --> 00:04:25,320 So this is Eversource down here, this is National Grid up there. 89 00:04:25,320 --> 00:04:26,970 And we don't know what's going on, 90 00:04:26,970 --> 00:04:30,060 whether there's a difference in its operating 91 00:04:30,060 --> 00:04:33,420 pressure of the pipes, or age of the pipes, 92 00:04:33,420 --> 00:04:36,840 material of the pipes, or difference in how the two 93 00:04:36,840 --> 00:04:39,200 companies deal with leaks. 94 00:04:39,200 --> 00:04:42,660 But you know, two years ago, when we did the survey, 95 00:04:42,660 --> 00:04:43,890 there were-- 96 00:04:43,890 --> 00:04:47,660 you know, this area of National Grid's was the Swiss Alps 97 00:04:47,660 --> 00:04:49,710 of Somerville. 98 00:04:49,710 --> 00:04:51,897 And it seems to still be that way. 99 00:04:51,897 --> 00:04:53,980 NATHAN PHILLIPS: So just a couple of observations. 100 00:04:53,980 --> 00:04:57,870 And please, you were there, so offer your own observations 101 00:04:57,870 --> 00:05:00,240 or questions about this because this 102 00:05:00,240 --> 00:05:02,030 is a very rich data set there. 103 00:05:02,030 --> 00:05:04,770 It raises lots of questions to me. 104 00:05:04,770 --> 00:05:07,650 So you can see here, there's actually a couple traces, 105 00:05:07,650 --> 00:05:08,360 right? 106 00:05:08,360 --> 00:05:09,960 Here there's two traces. 107 00:05:09,960 --> 00:05:12,540 We drove it two times in two trips. 108 00:05:12,540 --> 00:05:15,360 You see that there some variation each time. 109 00:05:15,360 --> 00:05:18,030 And we talked about the vagaries of wind, 110 00:05:18,030 --> 00:05:20,070 and you know, if the wind's blowing harder 111 00:05:20,070 --> 00:05:21,600 in one direction one time, you'll 112 00:05:21,600 --> 00:05:25,060 get something slightly different. 113 00:05:25,060 --> 00:05:26,030 So is this a leak? 114 00:05:26,030 --> 00:05:27,690 Is it the same thing as this? 115 00:05:27,690 --> 00:05:30,030 What do we actually-- 116 00:05:30,030 --> 00:05:33,820 how do we take continuous data like this 117 00:05:33,820 --> 00:05:38,640 and objectify it in terms of, well, there is a leak. 118 00:05:38,640 --> 00:05:41,520 It's difficult. It's tricky to do. 119 00:05:41,520 --> 00:05:43,800 And you know there are pipeline leaks. 120 00:05:43,800 --> 00:05:46,140 There are holes in pipes. 121 00:05:46,140 --> 00:05:48,210 And then there are methane leaks. 122 00:05:48,210 --> 00:05:50,370 And they come out of the ground. 123 00:05:50,370 --> 00:05:57,380 So even the terminology we use is not quite fully worked out. 124 00:05:57,380 --> 00:05:58,260 There's a lot here. 125 00:06:01,500 --> 00:06:02,730 Some of the things that-- 126 00:06:02,730 --> 00:06:06,660 so in terms of, at one level, at a very broad level, 127 00:06:06,660 --> 00:06:08,650 Pearl Street is a mess. 128 00:06:08,650 --> 00:06:12,450 It's leaking methane all along that street. 129 00:06:12,450 --> 00:06:15,270 When we get to policy, state policy 130 00:06:15,270 --> 00:06:17,280 makers and the utilities, and the regulators 131 00:06:17,280 --> 00:06:20,250 want to know, well, how many leaks are there? 132 00:06:20,250 --> 00:06:22,912 Because that determines, like, how many 133 00:06:22,912 --> 00:06:24,870 crews that are going to go out, and how they're 134 00:06:24,870 --> 00:06:26,828 going to schedule, and how they're going to get 135 00:06:26,828 --> 00:06:31,200 reimbursed, or repaid for a fixed number of things 136 00:06:31,200 --> 00:06:32,280 that they do. 137 00:06:32,280 --> 00:06:36,540 So this is a complicated kind of policy science 138 00:06:36,540 --> 00:06:39,772 type of framework. 139 00:06:39,772 --> 00:06:41,230 Do you have any other observations? 140 00:06:41,230 --> 00:06:42,980 AUDREY SCHULMAN: Yeah, I want to point out 141 00:06:42,980 --> 00:06:45,630 one thing in terms of policy. 142 00:06:45,630 --> 00:06:49,330 This area right here is where they'd 143 00:06:49,330 --> 00:06:53,340 taken out the cast iron main that is quite old 144 00:06:53,340 --> 00:06:54,690 and put in plastic. 145 00:06:54,690 --> 00:06:56,567 And we could see-- 146 00:06:56,567 --> 00:06:57,650 I think that that's right. 147 00:06:57,650 --> 00:06:58,070 NATHAN PHILLIPS: Franklin, is that Franklin? 148 00:06:58,070 --> 00:06:59,861 AUDREY SCHULMAN: Yeah I'm assuming that is. 149 00:06:59,861 --> 00:07:01,110 We can go in closer. 150 00:07:01,110 --> 00:07:03,390 But there was one part where we saw 151 00:07:03,390 --> 00:07:05,190 that they've taken out the old gas main 152 00:07:05,190 --> 00:07:07,650 and put in a new, better one. 153 00:07:07,650 --> 00:07:09,179 And I'm betting it's that area. 154 00:07:09,179 --> 00:07:10,887 AUDIENCE: So, were you surprised by this? 155 00:07:15,110 --> 00:07:17,750 NATHAN PHILLIPS: Not any more. 156 00:07:17,750 --> 00:07:24,750 So when we published our first publication in late 2012, 157 00:07:24,750 --> 00:07:30,120 early 2013, it basically made it a problem 158 00:07:30,120 --> 00:07:35,790 that was well known to the utilities known to, at least, 159 00:07:35,790 --> 00:07:40,310 I don't want to say everyone, but people who saw the paper, 160 00:07:40,310 --> 00:07:43,760 and members of the public that read some of the press that 161 00:07:43,760 --> 00:07:45,360 came out about it. 162 00:07:45,360 --> 00:07:50,930 So that was unexpected to the utilities. 163 00:07:50,930 --> 00:07:52,680 And it kind of knocked them back. 164 00:07:52,680 --> 00:07:55,990 It's like, oh, we have data that's coming out. 165 00:07:55,990 --> 00:07:59,210 So they didn't control the data. 166 00:07:59,210 --> 00:08:04,920 And so just by virtue of having data out there, 167 00:08:04,920 --> 00:08:08,960 it has basically, kind of say, balanced 168 00:08:08,960 --> 00:08:14,000 the power a little bit, or a lot, to get change to happen 169 00:08:14,000 --> 00:08:17,136 and policy to be made. 170 00:08:17,136 --> 00:08:17,890 Is this Franklin? 171 00:08:17,890 --> 00:08:18,930 AUDREY SCHULMAN: Yeah, I don't know. 172 00:08:18,930 --> 00:08:19,555 I couldn't see. 173 00:08:19,555 --> 00:08:22,960 I tried to figure it out. 174 00:08:22,960 --> 00:08:24,520 Anybody know Somerville really well? 175 00:08:24,520 --> 00:08:26,187 What's the street parallel to Roland? 176 00:08:26,187 --> 00:08:28,520 AUDIENCE: Can you assign these leaks to a specific pipe? 177 00:08:36,870 --> 00:08:42,250 What's the spatial resolution of assigning a spike to the--? 178 00:08:42,250 --> 00:08:47,860 NATHAN PHILLIPS: Yeah, so when we go by here, 179 00:08:47,860 --> 00:08:51,850 any one of these kind of discrete spikes, the window 180 00:08:51,850 --> 00:08:55,120 in which leaks may be coming out is probably 181 00:08:55,120 --> 00:08:57,715 on the border of a few tens of meters. 182 00:09:00,940 --> 00:09:05,740 And that's dictated both by the spatial source. 183 00:09:05,740 --> 00:09:08,090 It may-- even one pipeline leak, like I said, 184 00:09:08,090 --> 00:09:10,270 could be coming up in various locations, 185 00:09:10,270 --> 00:09:12,350 and then the wind may be blowing it around. 186 00:09:12,350 --> 00:09:14,170 So that's what this is allowing us to do. 187 00:09:14,170 --> 00:09:19,750 But then streets sometimes have multiple pipelines running down 188 00:09:19,750 --> 00:09:20,440 the same street. 189 00:09:20,440 --> 00:09:25,530 Sometimes they only have one main running down the street. 190 00:09:25,530 --> 00:09:29,260 So you have to go in a little more carefully 191 00:09:29,260 --> 00:09:33,370 and use, basically, probes that check 192 00:09:33,370 --> 00:09:36,760 what's coming out of the ground to actually pinpoint 193 00:09:36,760 --> 00:09:39,520 the actual pipe and where it's leaking. 194 00:09:39,520 --> 00:09:41,890 So this doesn't do this. 195 00:09:41,890 --> 00:09:45,640 This basically just says, Pearl Street is a mess, 196 00:09:45,640 --> 00:09:47,530 and it's got a lot of leak problems. 197 00:09:47,530 --> 00:09:52,265 And it requires a walking survey to go back and to be more-- 198 00:09:52,265 --> 00:09:54,640 AUDIENCE: Have you done some very basic signal processing 199 00:09:54,640 --> 00:09:55,140 on it? 200 00:09:57,610 --> 00:10:00,850 So run a low pass filter just to get rid 201 00:10:00,850 --> 00:10:03,810 of the small frequencies? 202 00:10:03,810 --> 00:10:07,390 NATHAN PHILLIPS: No, and I think that's the type of analysis 203 00:10:07,390 --> 00:10:10,120 that would be fabulous to do here. 204 00:10:10,120 --> 00:10:12,027 AUDREY SCHULMAN: Know anybody who can help? 205 00:10:12,027 --> 00:10:13,360 AUDIENCE: I can take it farther. 206 00:10:18,404 --> 00:10:19,320 NATHAN PHILLIPS: Yeah? 207 00:10:19,320 --> 00:10:30,190 AUDIENCE: [INAUDIBLE] You can detect the pattern that 208 00:10:30,190 --> 00:10:32,004 reflects the leak-- 209 00:10:32,004 --> 00:10:32,920 NATHAN PHILLIPS: Yeah. 210 00:10:32,920 --> 00:10:40,540 AUDIENCE: [INAUDIBLE] 211 00:10:40,540 --> 00:10:42,460 NATHAN PHILLIPS: So here's one thing 212 00:10:42,460 --> 00:10:46,020 that I think we all Audrey and I have 213 00:10:46,020 --> 00:10:51,280 been trying to make progress on is, how do you take this data 214 00:10:51,280 --> 00:10:56,070 and use it to help us start to quantify how much is coming 215 00:10:56,070 --> 00:10:56,830 out? 216 00:10:56,830 --> 00:11:00,050 And that's also a very difficult thing because, first of all, 217 00:11:00,050 --> 00:11:02,730 you're looking at this, and it implies something quantitative, 218 00:11:02,730 --> 00:11:04,480 but you don't see any numbers here, right? 219 00:11:04,480 --> 00:11:08,290 So probably you're wondering, well, how big are these spikes? 220 00:11:08,290 --> 00:11:11,260 And actually, you know, this data just came out, 221 00:11:11,260 --> 00:11:13,720 so I haven't been able to process, 222 00:11:13,720 --> 00:11:17,490 but we know that this is sitting right around 2.0, 223 00:11:17,490 --> 00:11:21,040 maybe 1.95, I forget what our baseline is. 224 00:11:21,040 --> 00:11:22,154 It's in the KML file. 225 00:11:22,154 --> 00:11:23,070 We could open that up. 226 00:11:23,070 --> 00:11:24,010 It's just a text file. 227 00:11:24,010 --> 00:11:27,040 It tells you what we set the baseline at-- 228 00:11:27,040 --> 00:11:30,100 that baseline can shift from day to day, by the way. 229 00:11:30,100 --> 00:11:31,010 What is this value? 230 00:11:31,010 --> 00:11:32,140 I'll have to find it out. 231 00:11:32,140 --> 00:11:33,520 It would be great to have-- 232 00:11:33,520 --> 00:11:36,340 we were talking about something that some students could 233 00:11:36,340 --> 00:11:39,614 do is come up with a processor that just basically plops 234 00:11:39,614 --> 00:11:40,280 numbers on here. 235 00:11:40,280 --> 00:11:41,560 So we could see that. 236 00:11:41,560 --> 00:11:43,480 What was the concentration? 237 00:11:43,480 --> 00:11:46,450 If we go, you know, just for setting the scale, 238 00:11:46,450 --> 00:11:49,480 I remember, and the four of us in the van 239 00:11:49,480 --> 00:11:53,890 for the second round, we went to Sullivan Square. 240 00:11:53,890 --> 00:11:56,530 And I think, what was the top read we got there? 241 00:11:56,530 --> 00:11:57,519 We were reading it off. 242 00:11:57,519 --> 00:11:59,310 AUDREY SCHULMAN: It was like 90, wasn't it? 243 00:11:59,310 --> 00:12:00,850 Wasn't it something really ridiculous? 244 00:12:00,850 --> 00:12:02,308 I've never seen that number before. 245 00:12:02,308 --> 00:12:06,250 NATHAN PHILLIPS: I recall up around 67 or 70 parts 246 00:12:06,250 --> 00:12:08,270 per million in the air. 247 00:12:08,270 --> 00:12:12,017 AUDREY SCHULMAN: So it's that spike right there, see? 248 00:12:12,017 --> 00:12:14,350 NATHAN PHILLIPS: So there's something really interesting 249 00:12:14,350 --> 00:12:14,990 about this. 250 00:12:14,990 --> 00:12:16,720 So we can put numbers on that. 251 00:12:16,720 --> 00:12:19,030 We have done that, and it's just, 252 00:12:19,030 --> 00:12:22,150 we haven't had a chance to do it. 253 00:12:22,150 --> 00:12:27,430 But actually this is really a very interesting issue 254 00:12:27,430 --> 00:12:29,080 of data display. 255 00:12:29,080 --> 00:12:31,510 I'd be interested-- who's that data person that, 256 00:12:31,510 --> 00:12:33,964 the display of data-- 257 00:12:33,964 --> 00:12:34,880 AUDIENCE: [INAUDIBLE]. 258 00:12:34,880 --> 00:12:37,213 NATHAN PHILLIPS: Yeah, I'd be interested to hear someone 259 00:12:37,213 --> 00:12:39,340 like that's take because there was a utility 260 00:12:39,340 --> 00:12:42,520 person from National Grid that basically called us out 261 00:12:42,520 --> 00:12:44,480 on this. 262 00:12:44,480 --> 00:12:46,480 And the point this person made was, 263 00:12:46,480 --> 00:12:49,650 it's like, you're scaring people by doing this. 264 00:12:49,650 --> 00:12:52,780 You're taking-- it's an apples and oranges thing. 265 00:12:52,780 --> 00:12:55,000 You're putting some data. 266 00:12:55,000 --> 00:12:58,690 And all of the houses, and all of us, are, like, down here, 267 00:12:58,690 --> 00:13:01,750 and you're creating these things that are just making it 268 00:13:01,750 --> 00:13:05,110 look like the world is ending. 269 00:13:05,110 --> 00:13:08,920 And I kind of get that, at one level, 270 00:13:08,920 --> 00:13:13,690 is that this is really apples to oranges type of thing 271 00:13:13,690 --> 00:13:18,340 because we're conflating PPM values parts, per million 272 00:13:18,340 --> 00:13:21,480 with meters, or space. 273 00:13:21,480 --> 00:13:23,390 And they are two different beasts, 274 00:13:23,390 --> 00:13:25,090 and we're like putting them together. 275 00:13:25,090 --> 00:13:26,320 But I will say this. 276 00:13:26,320 --> 00:13:27,790 As a scientist, the first thing I 277 00:13:27,790 --> 00:13:31,780 was trained as a freshman was, if you have data, 278 00:13:31,780 --> 00:13:33,850 and you're going to graph it, use 279 00:13:33,850 --> 00:13:37,870 the space available to fit your data, you know, 280 00:13:37,870 --> 00:13:39,080 and maximize that space. 281 00:13:39,080 --> 00:13:41,420 So when I want to plot this, I want 282 00:13:41,420 --> 00:13:44,800 to make it visible as best I can. 283 00:13:44,800 --> 00:13:46,969 AUDIENCE: [INAUDIBLE] I think it might 284 00:13:46,969 --> 00:13:48,552 be useful to come up with a unit, that 285 00:13:48,552 --> 00:13:53,870 is understandable as a measurement unit. 286 00:13:53,870 --> 00:13:56,100 It could be anything. 287 00:13:56,100 --> 00:13:58,970 And I will not mention that word, 288 00:13:58,970 --> 00:14:03,170 but stuff that comes out of us as-- 289 00:14:03,170 --> 00:14:06,900 I say, you know, what is the concept? 290 00:14:06,900 --> 00:14:09,046 So something that says-- 291 00:14:11,070 --> 00:14:12,460 AUDREY SCHULMAN: Nathan's idea is 292 00:14:12,460 --> 00:14:15,770 to sort of take just sort of some number that 293 00:14:15,770 --> 00:14:18,440 covers this whole area for this leak, 294 00:14:18,440 --> 00:14:21,315 so that it adds up the whole area underneath that. 295 00:14:21,315 --> 00:14:22,830 AUDIENCE: So you can bin that. 296 00:14:22,830 --> 00:14:27,050 But what I'm saying, as an understandable measure, 297 00:14:27,050 --> 00:14:32,930 something that says the height is measured in terms of a unit 298 00:14:32,930 --> 00:14:38,390 that we all recognize as a relevant unit for gas. 299 00:14:38,390 --> 00:14:42,310 [INAUDIBLE] 300 00:14:42,310 --> 00:14:45,830 AUDIENCE: But you could be able to reverse model it, right? 301 00:14:45,830 --> 00:14:47,980 So put it in an atmospheric model 302 00:14:47,980 --> 00:14:51,650 and say the concentration at three feet above the ground 303 00:14:51,650 --> 00:14:54,030 is X. What's the leak rate have to be 304 00:14:54,030 --> 00:14:58,430 to get that concentration here in this pattern? 305 00:14:58,430 --> 00:15:00,827 And figure out, how much they're actually emitting, 306 00:15:00,827 --> 00:15:03,630 and then you have a number that is at least in terms 307 00:15:03,630 --> 00:15:04,977 of their dollars, right? 308 00:15:04,977 --> 00:15:07,060 AUDREY SCHULMAN: But it's hard to do that somewhat 309 00:15:07,060 --> 00:15:08,776 because the wind will be going by. 310 00:15:08,776 --> 00:15:10,317 NATHAN PHILLIPS: Right, but the wind, 311 00:15:10,317 --> 00:15:12,275 so you do it on a couple of days, and you have, 312 00:15:12,275 --> 00:15:15,430 you put in the wind data to your model. 313 00:15:15,430 --> 00:15:19,390 And then your model says, this methane at this first rate, 314 00:15:19,390 --> 00:15:21,680 it back-calculates-- 315 00:15:21,680 --> 00:15:24,800 from your concentration, back-calculates an emission 316 00:15:24,800 --> 00:15:26,850 based on the weather data as well. 317 00:15:26,850 --> 00:15:29,675 Normalize that over multiple days of testing, 318 00:15:29,675 --> 00:15:32,780 and then you say, this roughly equates to X emission. 319 00:15:32,780 --> 00:15:35,600 And there are research groups that 320 00:15:35,600 --> 00:15:38,100 are trying to do exactly that. 321 00:15:38,100 --> 00:15:41,600 So what you have are the turbulence experts, the people 322 00:15:41,600 --> 00:15:44,990 that study micro meteorology, and boundary layers, 323 00:15:44,990 --> 00:15:46,130 and I'm not that person. 324 00:15:46,130 --> 00:15:47,465 I know that person. 325 00:15:47,465 --> 00:15:49,410 AUDREY SCHULMAN: I don't even know dispersion. 326 00:15:49,410 --> 00:15:54,070 NATHAN PHILLIPS: But these are the future areas to take this 327 00:15:54,070 --> 00:15:55,820 and to try to model. 328 00:15:55,820 --> 00:15:57,861 AUDIENCE: See, I'm wondering against whether this 329 00:15:57,861 --> 00:16:00,110 is a physics problem or a statistics problem. 330 00:16:00,110 --> 00:16:01,820 AUDREY SCHULMAN: Yes. 331 00:16:01,820 --> 00:16:04,016 NATHAN PHILLIPS: Yes, it is. 332 00:16:04,016 --> 00:16:06,140 AUDIENCE: Because I feel like some of these things, 333 00:16:06,140 --> 00:16:09,020 you could get very complex physics because 334 00:16:09,020 --> 00:16:11,710 of the nature of the scales at which you 335 00:16:11,710 --> 00:16:16,450 were trying to model this thing, or run a machine learning 336 00:16:16,450 --> 00:16:19,050 algorithm against it and see what is your best predictor. 337 00:16:19,050 --> 00:16:20,716 NATHAN PHILLIPS: And I just want to make 338 00:16:20,716 --> 00:16:23,300 a really quick point here is, like here's 339 00:16:23,300 --> 00:16:25,160 one of these spikes, right? 340 00:16:25,160 --> 00:16:29,030 This value, let's say it's 10 PPM, that's 341 00:16:29,030 --> 00:16:34,490 important on its own because, for example, air quality is 342 00:16:34,490 --> 00:16:37,340 related to concentration of methane in the air. 343 00:16:37,340 --> 00:16:40,450 It's a precursor to ozone. 344 00:16:40,450 --> 00:16:44,480 The area under this curve, we think, 345 00:16:44,480 --> 00:16:49,490 may be a correlant for how much is coming out. 346 00:16:49,490 --> 00:16:51,820 And that becomes an energy point, 347 00:16:51,820 --> 00:16:57,180 or an equivalent carbon unit is this area under the curve. 348 00:16:57,180 --> 00:16:59,660 But you know, these need to be validated. 349 00:16:59,660 --> 00:17:01,640 Everyone in the class now, I think, 350 00:17:01,640 --> 00:17:05,599 I sent the email out, you have the KML file. 351 00:17:05,599 --> 00:17:07,829 You have this data. 352 00:17:07,829 --> 00:17:11,609 You have an ASCII data file, which generated this. 353 00:17:11,609 --> 00:17:14,819 You'd have to-- the columns are actually 354 00:17:14,819 --> 00:17:17,467 pretty easy understand, for the most part. 355 00:17:17,467 --> 00:17:18,550 But you are now empowered. 356 00:17:18,550 --> 00:17:19,633 This is community science. 357 00:17:19,633 --> 00:17:22,400 And you can explore, and analyze, 358 00:17:22,400 --> 00:17:24,180 and run with this data. 359 00:17:24,180 --> 00:17:27,826 AUDIENCE: I could forward it to everybody who's signed in. 360 00:17:27,826 --> 00:17:29,490 Nathan, I have an unrelated question 361 00:17:29,490 --> 00:17:29,740 that I'm really curious about. 362 00:17:29,740 --> 00:17:31,823 Have you ever driven near a cow farm, a beef farm? 363 00:17:36,620 --> 00:17:40,280 NATHAN PHILLIPS: So the one that sticks out in my mind is 364 00:17:40,280 --> 00:17:46,850 driving from San Francisco to LA on Interstate 5 getting close 365 00:17:46,850 --> 00:17:52,550 to the Southern Mountains, giant feed lot right next to I-5, 366 00:17:52,550 --> 00:17:55,730 and very flat baseline, and then, you know, 367 00:17:55,730 --> 00:17:58,910 this very sloping increase as we passed by that feed lot 368 00:17:58,910 --> 00:18:02,390 to about, I think it was 2 and 1/2, 369 00:18:02,390 --> 00:18:06,610 3 parts per million of methane going from below 2-- 370 00:18:06,610 --> 00:18:09,382 but the thing there, it's not like this spike 371 00:18:09,382 --> 00:18:10,340 like we're seeing here. 372 00:18:10,340 --> 00:18:12,240 It's just this blob. 373 00:18:12,240 --> 00:18:13,670 It's a very extended blob. 374 00:18:13,670 --> 00:18:15,920 AUDIENCE: That's why I think integration is important. 375 00:18:15,920 --> 00:18:21,060 So you want to do not just a spike but a spike times area. 376 00:18:24,060 --> 00:18:26,580 AUDIENCE: Which an air dispersion model 377 00:18:26,580 --> 00:18:30,021 would essentially-- would do. 378 00:18:30,021 --> 00:18:32,020 AUDREY SCHULMAN: So we're going to need somebody 379 00:18:32,020 --> 00:18:33,368 to help us with this. 380 00:18:33,368 --> 00:18:35,243 NATHAN PHILLIPS: Yeah, these are great ideas. 381 00:18:35,243 --> 00:18:35,743 Question? 382 00:18:35,743 --> 00:18:38,630 AUDIENCE: I just wanted to know, can you 383 00:18:38,630 --> 00:18:41,590 get ahold of this device? 384 00:18:41,590 --> 00:18:43,950 How much does it cost? 385 00:18:43,950 --> 00:18:46,190 NATHAN PHILLIPS: The Picarro analyzer that we used-- 386 00:18:46,190 --> 00:18:51,340 and that's now six, seven years old, that was about $60,000. 387 00:18:51,340 --> 00:18:54,660 And the manufacturer, it's kind of like, 388 00:18:54,660 --> 00:18:55,950 they don't have a price list. 389 00:18:55,950 --> 00:18:57,510 They're like, call us and let's talk. 390 00:18:57,510 --> 00:18:59,010 AUDIENCE: So that's just for the box 391 00:18:59,010 --> 00:19:01,714 with the mirrors and the laser. 392 00:19:01,714 --> 00:19:02,880 NATHAN PHILLIPS: Yeah, yeah. 393 00:19:02,880 --> 00:19:04,035 So you know, it's not really-- 394 00:19:04,035 --> 00:19:06,451 AUDIENCE: This is the kind of stuff we have to [INAUDIBLE] 395 00:19:06,451 --> 00:19:07,380 NATHAN PHILLIPS: Yeah. 396 00:19:07,380 --> 00:19:10,260 But, what I will say is that, or maybe 397 00:19:10,260 --> 00:19:13,785 Audrey wants to talk about the hand held? 398 00:19:13,785 --> 00:19:15,802 The Sierra Club HEET thing. 399 00:19:15,802 --> 00:19:17,760 AUDREY SCHULMAN: Yeah HEET might, at some point 400 00:19:17,760 --> 00:19:20,160 have a through Sierra Club, Massachusetts 401 00:19:20,160 --> 00:19:24,362 have a handheld device with which you could check something 402 00:19:24,362 --> 00:19:26,486 like that combustible gas indicator for anybody who 403 00:19:26,486 --> 00:19:28,890 is on the ride yesterday. 404 00:19:28,890 --> 00:19:32,185 So you'd be able to check the gas in the soil 405 00:19:32,185 --> 00:19:37,020 to find out if a tree is being poisoned by gas-- et cetera. 406 00:19:37,020 --> 00:19:39,650 NATHAN PHILLIPS: And I can't forget-- someone, who was it, 407 00:19:39,650 --> 00:19:42,540 came up with the idea of a bike, a bike trailer. 408 00:19:42,540 --> 00:19:43,290 AUDIENCE: You did. 409 00:19:43,290 --> 00:19:44,835 That was your idea. 410 00:19:44,835 --> 00:19:45,710 NATHAN PHILLIPS: Huh? 411 00:19:45,710 --> 00:19:46,349 AUDIENCE: You were talking about it. 412 00:19:46,349 --> 00:19:46,820 That was your idea. 413 00:19:46,820 --> 00:19:47,080 NATHAN PHILLIPS: Oh. 414 00:19:47,080 --> 00:19:47,579 Well-- 415 00:19:47,579 --> 00:19:53,300 [LAUGHTER] So that would be amazing. 416 00:19:53,300 --> 00:19:55,640 And there's a reason for it. 417 00:19:55,640 --> 00:19:58,730 Because, as you mentioned, you know these mains, 418 00:19:58,730 --> 00:20:01,740 they can run, more times than not, 419 00:20:01,740 --> 00:20:03,830 they run kind of down the middle of the street. 420 00:20:03,830 --> 00:20:05,360 But they can go on sidewalks. 421 00:20:05,360 --> 00:20:07,240 They can go on angles. 422 00:20:07,240 --> 00:20:08,890 The service lines leak. 423 00:20:08,890 --> 00:20:11,285 And so much of what we've done has not 424 00:20:11,285 --> 00:20:13,660 been looking at service lines, the perpendicular, smaller 425 00:20:13,660 --> 00:20:16,130 pipes that go into the houses. 426 00:20:16,130 --> 00:20:20,890 So with a bike or a cart, or an ability to get on sidewalks-- 427 00:20:20,890 --> 00:20:24,680 and to be able to move around in Boston or Cambridge, 428 00:20:24,680 --> 00:20:27,397 you know, if you ride a bike, that's the most efficient way 429 00:20:27,397 --> 00:20:27,980 to get around. 430 00:20:27,980 --> 00:20:30,450 I mean, if there's a gridlock, you're just going around. 431 00:20:30,450 --> 00:20:30,950 So-- 432 00:20:30,950 --> 00:20:32,180 AUDIENCE: Baby carriages. 433 00:20:32,180 --> 00:20:34,410 NATHAN PHILLIPS: Or baby carriages. 434 00:20:34,410 --> 00:20:39,430 And so Picarro does sell a smaller unit than that one, 435 00:20:39,430 --> 00:20:40,520 it's like a backpack unit. 436 00:20:40,520 --> 00:20:43,225 But it could go in a Burly trailer, 437 00:20:43,225 --> 00:20:46,670 or it could go on your back while you're riding a bike. 438 00:20:46,670 --> 00:20:47,694 So I would love that. 439 00:20:47,694 --> 00:20:49,360 It would be the first one in the nation, 440 00:20:49,360 --> 00:20:50,920 and I think it would be amazing. 441 00:20:50,920 --> 00:20:53,520 So if we can crowdsource that, generate some funds, 442 00:20:53,520 --> 00:20:54,515 that would be great. 443 00:20:54,515 --> 00:20:56,389 AUDIENCE: Well, couldn't the existing Picarro 444 00:20:56,389 --> 00:20:59,810 that you had in the van, can that be mounted, 445 00:20:59,810 --> 00:21:02,817 that and the battery or power, could that 446 00:21:02,817 --> 00:21:05,090 be mounted on our trailer behind a bike? 447 00:21:05,090 --> 00:21:06,520 NATHAN PHILLIPS: Yeah, it could. 448 00:21:06,520 --> 00:21:11,150 We could actually make a video with that, 449 00:21:11,150 --> 00:21:15,590 and that would be like, let's do it better-- 450 00:21:15,590 --> 00:21:19,000 because what you would use there is, like a garden cart, 451 00:21:19,000 --> 00:21:21,530 with those big balloon knobby tires 452 00:21:21,530 --> 00:21:26,244 to provide some shock absorption. 453 00:21:30,330 --> 00:21:31,830 AUDIENCE: Yeah, actually let me make 454 00:21:31,830 --> 00:21:35,180 a point of some order, which is now that we-- 455 00:21:35,180 --> 00:21:37,590 I think the formal portion of your presentation 456 00:21:37,590 --> 00:21:39,530 is done, right? 457 00:21:39,530 --> 00:21:40,530 AUDREY SCHULMAN: Right. 458 00:21:40,530 --> 00:21:42,030 AUDIENCE: So I think that we can now 459 00:21:42,030 --> 00:21:44,107 ask questions of all the speakers, not just-- 460 00:21:44,107 --> 00:21:45,940 NATHAN PHILLIPS: After giving Audrey a hand. 461 00:21:45,940 --> 00:21:50,820 [APPLAUSE] 462 00:21:50,820 --> 00:21:52,370 Can we have Susan's last question? 463 00:21:52,370 --> 00:21:55,785 AUDIENCE: I just was wondering because, on the field trip, 464 00:21:55,785 --> 00:21:58,510 I learned what the symbolism is by the gas company, 465 00:21:58,510 --> 00:22:01,000 and what kind of cast iron, it was written on the sidewalk 466 00:22:01,000 --> 00:22:01,500 here. 467 00:22:04,620 --> 00:22:11,550 If the company has the infrastructure in its archives, 468 00:22:11,550 --> 00:22:15,250 is there enough of a relationship 469 00:22:15,250 --> 00:22:19,130 between the age of the pipe and the material to know the leak, 470 00:22:19,130 --> 00:22:22,550 so you could make big assumptions about, OK, 471 00:22:22,550 --> 00:22:26,927 if you know that, then you don even have to do the monitoring. 472 00:22:26,927 --> 00:22:29,510 AUDREY SCHULMAN: The Department of Environmental Protection is 473 00:22:29,510 --> 00:22:32,140 making that exact point, in terms of basing all 474 00:22:32,140 --> 00:22:36,650 of the state-wide greenhouse gas methane emissions based 475 00:22:36,650 --> 00:22:39,850 on exactly that calibration-- miles of cast iron, 476 00:22:39,850 --> 00:22:44,710 miles of bare steel, et cetera, and making an assumption that-- 477 00:22:44,710 --> 00:22:49,270 they know the exact rate of emissions per mile of cast iron 478 00:22:49,270 --> 00:22:52,040 main, and that they don't even have 479 00:22:52,040 --> 00:22:56,500 to check, either top down, or bottom up, either. 480 00:22:56,500 --> 00:23:00,310 And I think that's-- 481 00:23:00,310 --> 00:23:05,350 you could make a guess at that, but you 482 00:23:05,350 --> 00:23:10,560 have to check to make sure it's right. 483 00:23:10,560 --> 00:23:12,810 NATHAN PHILLIPS: Yeah, they call the emissions factors 484 00:23:12,810 --> 00:23:14,350 and activity factors. 485 00:23:14,350 --> 00:23:17,350 And multiply A by B and get a leak rate. 486 00:23:17,350 --> 00:23:20,670 And these, I call them fudge factors, they basically say, 487 00:23:20,670 --> 00:23:23,860 cast iron has this many leaks per linear mile. 488 00:23:23,860 --> 00:23:25,456 That's your emissions factor. 489 00:23:25,456 --> 00:23:26,830 Well, how many miles do you have? 490 00:23:26,830 --> 00:23:29,551 That's your activity factor, multiply a by d, 491 00:23:29,551 --> 00:23:30,550 there is your leak rate. 492 00:23:30,550 --> 00:23:34,330 But those are based on very old-- 493 00:23:34,330 --> 00:23:36,835 not just old data, but very sparse data. 494 00:23:36,835 --> 00:23:39,960 AUDIENCE: But can you calibrate that with you leak data? 495 00:23:39,960 --> 00:23:41,710 NATHAN PHILLIPS: Yeah, that's we're doing. 496 00:23:44,191 --> 00:23:45,940 AUDREY SCHULMAN: I don't think it has much 497 00:23:45,940 --> 00:23:48,360 to do with each other. 498 00:23:48,360 --> 00:23:50,870 The Department of Environmental Protection emissions factors 499 00:23:50,870 --> 00:23:54,340 make it look like-- the problem is solved! 500 00:23:54,340 --> 00:23:56,320 And I don't think that that's true. 501 00:23:56,320 --> 00:23:59,410 AUDIENCE: There are [INAUDIBLE] emissions factors, right? 502 00:23:59,410 --> 00:24:03,790 So they are from the industry, and they're way outdated. 503 00:24:03,790 --> 00:24:06,619 Like, this is true, they're outdated for refineries, 504 00:24:06,619 --> 00:24:08,910 so when they go out and do modern testing on refineries 505 00:24:08,910 --> 00:24:11,022 they see the tanks leak at much higher rates 506 00:24:11,022 --> 00:24:12,730 than their emissions factors account for. 507 00:24:12,730 --> 00:24:15,340 NATHAN PHILLIPS: So there's a really interesting point 508 00:24:15,340 --> 00:24:17,750 the Commonwealth of Massachusetts 509 00:24:17,750 --> 00:24:21,142 apparently has done an enormous job at cleaning up our methane 510 00:24:21,142 --> 00:24:23,400 leak problem because, what they did was, 511 00:24:23,400 --> 00:24:29,470 the DEP utilized an early set of emissions factors natural gas 512 00:24:29,470 --> 00:24:33,940 leakage from the mid 90s, and then one that 513 00:24:33,940 --> 00:24:38,590 was much lower, from 2015 publication, 514 00:24:38,590 --> 00:24:42,930 and they linearly interpolated an emissions factor, so that-- 515 00:24:42,930 --> 00:24:46,437 they hardwired in a reduction because two studies 516 00:24:46,437 --> 00:24:48,020 showed two different emissions factors 517 00:24:48,020 --> 00:24:49,730 for the same kind of pipe. 518 00:24:49,730 --> 00:24:52,360 So result you get, it appears to make it look like we've 519 00:24:52,360 --> 00:24:55,464 done really, really well. 520 00:24:55,464 --> 00:24:57,630 AUDREY SCHULMAN: Kind of like, everything is solved. 521 00:24:57,630 --> 00:25:00,170 And part of the way they did they was, they got rid 522 00:25:00,170 --> 00:25:06,734 of superemitters, or one of the studies discarded any outliers. 523 00:25:06,734 --> 00:25:08,400 NATHAN PHILLIPS: Right, so the emissions 524 00:25:08,400 --> 00:25:14,280 factors are based on leaks are distributed like this. 525 00:25:14,280 --> 00:25:19,050 But what we know is that leaks are distributed like that. 526 00:25:19,050 --> 00:25:21,390 They're long tail with a few-- 527 00:25:21,390 --> 00:25:23,540 AUDIENCE: I mean, I would say that the normal 528 00:25:23,540 --> 00:25:30,099 versus long tail is an error across so many problems 529 00:25:30,099 --> 00:25:31,206 and industries. 530 00:25:31,206 --> 00:25:33,167 NATHAN PHILLIPS: Yes. 531 00:25:33,167 --> 00:25:35,500 AUDREY SCHULMAN: And we can all take a look at this data 532 00:25:35,500 --> 00:25:36,666 and make a guess as to where 533 00:25:36,666 --> 00:25:38,940 the long tail problem is. 534 00:25:38,940 --> 00:25:40,504 It's fairly apparent. 535 00:25:40,504 --> 00:25:43,345 AUDIENCE: So it's 50% from 7%? 536 00:25:43,345 --> 00:25:44,720 NATHAN PHILLIPS: That's what we-- 537 00:25:44,720 --> 00:25:46,840 Margaret Hendrick did a study-- 538 00:25:46,840 --> 00:25:51,360 that we did together-- that 7 of 100 leaks 539 00:25:51,360 --> 00:25:55,390 accounted for 50% of the gas loss. 540 00:25:55,390 --> 00:25:58,500 AUDIENCE: If you take this data as kind of the sample, 541 00:25:58,500 --> 00:26:01,560 and multiply by the number of rows and pipelines, 542 00:26:01,560 --> 00:26:04,445 and you try to match it with entire methane 543 00:26:04,445 --> 00:26:08,150 inventory of the US, just to see if the numbers are 544 00:26:08,150 --> 00:26:09,400 the same orders of magnitude-- 545 00:26:09,400 --> 00:26:10,650 NATHAN PHILLIPS: Yeah, we did. 546 00:26:10,650 --> 00:26:14,660 Not with the entire US, but we did it with Massachusetts 547 00:26:14,660 --> 00:26:17,300 AUDIENCE: [INAUDIBLE] 548 00:26:17,300 --> 00:26:21,000 NATHAN PHILLIPS: Yeah, so we found 549 00:26:21,000 --> 00:26:24,750 that, if we took the chamber measurements of leaks, 550 00:26:24,750 --> 00:26:27,120 100 leaks, and we just kind of multiplied that out 551 00:26:27,120 --> 00:26:32,060 by the frequency of leaks that we got previously-- 552 00:26:32,060 --> 00:26:37,540 that amount is consistent with other estimates, 553 00:26:37,540 --> 00:26:43,080 and are about one third of the total methane 554 00:26:43,080 --> 00:26:47,840 emissions estimated for the Commonwealth of Massachusetts. 555 00:26:47,840 --> 00:26:50,240 One third. 556 00:26:50,240 --> 00:26:55,950 This collaboration, we have to me, and with Mothers Out Front 557 00:26:55,950 --> 00:27:01,170 as a kind of hybrid advocacy science 558 00:27:01,170 --> 00:27:04,710 nonprofit kind of coalition, I've 559 00:27:04,710 --> 00:27:07,470 never been involved in anything like that, 560 00:27:07,470 --> 00:27:10,670 and it's just been the most fulfilling kind of partnership 561 00:27:10,670 --> 00:27:12,560 for research science and policy. 562 00:27:12,560 --> 00:27:14,110 AUDREY SCHULMAN: Yeah.